Skip to main content

Full text of "The American journal of science"

See other formats


,5^0' 73 



THE 

AMERICAN 

JOURNAL OF SCIENCE. 

Editor: EDWARD S. DANA. 

ASSOCIATE EDITORS 

Professors GEOKGE L. GOODALE, JOHN TEOWBKIDGE, 
W. G. FARLOW and WM. M. DAVIS, of Cambridge, 

Professors ADDISON E. VERRILL, HORACE L. WELLS, 
L. Y. PIRSSON and H. E. GREGORY, of New Haven, 

Professor GEORGE F. BARKER, of Philadelphia, 
Professor HENRY S. WILLIAMS, of Ithaca, 
Professor JOSEPH S. AMES, of Baltimore, 
Mr. J. S. DILLER, of Washington. 



FOURTH SERIES 

VOL. XX VII— [WHOLE NUMBER, CLXXVIL] 

WITH PLATES I-IV. 



NEW HAVEN, CONNECTICUT. 

1909. 



4oG£3o 



THE TUTTLE, MOREHOUSE & TAYLOR COMPANY, 
NEW HAVEN. 



CONTENTS TO VOLUME XXVII. 



N^ ber 157. 

Page 
Art. I. — Diopside and its Relations to Calcium and Magne- 
sium Metasilicates ; by E. T. Allen and VV. P. White. 
With Optical Study ; by F. E. Wright and E. S. 

Larsen. (With Plate I) 1 

II— California Earthquake of 1906; by G-. K. Gilbert 48 

III— Descriptions of Tertiary Insects ; by T. D. A. Cock- 

erell. Part V __. 53 

IV. — Electrolytic Estimation of Lead and of Manganese by 
the Use of the Filtering Crucible ; by F. A. Gooch and 

F. B. Beyer 59 

V. — Specific Radio-Activity of Thorium and its Products ; 

by G. C. Ashman . 65 

VI. — Coronas with Mercury Light ; by C. Barus 73 

SCIENTIFIC INTELLIGENCE. 

Chemistry mid Physics— Attempt to Produce a Compound of Argon, Fischer 
and Iliovici : Explosive Crystallization, Weston, 82. — Constituents of 
Ytterbium, A. v. Welsbach : New Form of " Tin Infection," von Hass- 
linger : Heat Evolved by Radium, von Schweidler and Hess, 83. — 
Positive Rays, W. Wien : Speclral Intensity of Canal Rays, J. Stark and 
W. Stenberg : Canal Rays, J. Stark : Potential Measurements in the 
dark Cathode Space, W. Westphal, 84. — Elements of Physics, Nichols 
and Franklin : Text-Book of Physics, 85. 

Geology — Publications of the United States Geological Survey, 86. — Canada 
Geological Survey : North Carolina Geological and Economic Survey, 
87.— Report of the State Geologist of Vermont for 1907-8 : Thirty- 
second Annual Report, Indiana Department of Geology, 88. — Illinois 
State Geological Survey, Year-Book for 1907 : New Zealand Geolog- 
ical Survey : Report on the Eruptions of the Soufriere in St. Vincent in 
1902. and on a Visit to Montagne Pelee in Martinique, T. Anderson and 
J. S. Flett, 89. —Geology and Ore deposits of the Coeur d' Alene District, 
Idaho, 90. — Geologie der Steinkohlenlager, Dannenberg : Geology of 
Coal and Coal-mining, W. Gibson : Physical History of the Earth in Out- 
line, J. B. Babbitt: Triassic Ichthyosauria. with special reference to 
American Forms, J. C. Merriam, 91. — Fossil Vertebrates in the American 
Museum of Natural History, Part I — Fishes, L. Httssakof, 92. — Conrad 
Fissure, B. Brown : Four-horned Pelycosaurian from the Permian of 
Texas, W. D. Matthew : Osteology of Blastomeryx and Phylogeny of the 
American Cervidae, W. D. Matthew : Rhinoceroses from the Oligocene 
and Miocene deposits of North Dakota and Montana, E. Douglass, 93. — 
Fossil Horses from North Dakota and Montana, E. Douglass : Some Oli- 
gocene Lizards, E. -Douglass : Preliminary Notes on Some American 
Chalicotheres, 0. A. Peterson, 94. 

Botany and Zoology — Harvard Botanical Station in Cuba, 94. — Handbuch 
der Bluten-biologie, P. Knuth : Convenient Clearing and Mounting Agent, 
96. — Economic Zoology : Text-book of the Principles of Animal Histology, 
IT. Dahlgren and W. A. Kepner : Archiv. fur Zellforschung, R. Gold- 
schmidt : Ueber die Eibildung bei der Milbe Pediculopsis graminum, 97. 

Miscellaneous Scientific Intelligence — Artificial Daylight for Use with the 
Microscope, F. E. Wright : Ion ; A Journal of Electrotonics, Atomistics, 
Ionology, Radio-activity and Raumchemistry, 98. — Life and Letters of 
Herbert Spencer, D. Duncan, 99. — American Association for the Advance- 
ment of Science : Nature of Enzyme Action : Rivista di Scienza, 100. 

Obituary— 0. W. Gibbs : W. E. Ayrton, 100. 



IV CONTENTS. 



Number 158. 

Page 
Art. VII. — Revision of the Protostegidae ; by G. R. Wieland. 

(With Plates II-IY) 101 

VIII. — Submarine Eruptions of 1831 and 1891 near Pantel- 

leria ; by H. S. Washington _ _ _ . 131 

IX. — Types of Permian Insects; by E. H. Sellards .._ 151 

X. — Iodometric Estimation of Vanadic Acid, Chromic Acid 

and Iron in the Presence of One Another ; by G. Edgar 174 

XI. — Analysis and Chemical Composition of the Mineral 

War wickite ; by W. M. Bradley . _ 179 

SCIENTIFIC INTELLIGENCE. 

Chemistry and Physics — Question of Change in Total Weight of Chemically 
Eeacting Substances, H. Landolt : Volume of Eadium Emanation, 
Eutherford, 185. — New Method for Separating Tungstic and Silicic 
Oxides, Defacqz : Silicide of Uranium, Defacqz : New Periodic Func- 
tion of the Atomic Weight, V. Poschl, 186. — Velocity of Eontgen Eays ; 
also their Influence on the Brush Discharge, E. Marx : Eadiation of 
Uranium X, H. W. Schmidt : Influence of Self Induction on Spark 
Spectra, G-. Berndt : Ionization of Gases by Spark and Arc, H. Eausch, 
187. — Investigations in Eadiation, W. W. Coblentz and P. G. Nutting, 
18& 

Geology — Twenty-ninth Annual Eeport United States Geological Survey, 
G. 0. Smith, 188. — Geological Survey of New Jersey, H. B. Kummel : 
Sketch of the Geography and Geology of the Himalaya Mountains and 
Tibet, S. G. Burrard and H. H. Hayden, 189.— Gases in Eocks, E. T. 
Chamberlin, 190. 

Botany and Zoology — Forest Flora of New South Wales, J. H. Maiden, 
191 — Jaarboek van het Department van Langbouw in Niederlandsch- 
Indie, 1907: Origin of Vertebrates, W. H. Gaskell. 192.— Ticks: a 
Monograph of the Ixodoidea, G. H. F. Nuttall, C. Warburton, W. 
F. Cooper, and L. E. Eobinson : Animal Eomances, G. Eenshaw : Essays 
on Evolution, E. B. Poulton, 193. — Parasitology, G. H. F. Nutall and 
A. E. Shipley, 194. 

Miscellaneous Scientific Intelligence — New Goniometer Lamp, F. E. Wright, 
194. — Containing Device for Salts Used as Sources for Monochromatic 
Light, F. E. Wright, 195. — Eeport of the Secretary of the Smithsonian 
Institution for the year ending June 30, 1908, 196. 

Obituary— G. W. Hough, 196. 



CONTENTS. 



Number 159. 

Page 

Art. XII. — Recent Observations in Atmospheric Electricity; 

by P. H. Dike 197 

XIII. — Iodyrite from Tonopah, Nevada, and Broken Hill, 

New South Wales ; by E. H. Kraus and C. W. Cook__ 210 

XIV. — Deviation of Rays by Prisms; by H. S. Uhler 223 

XV. — Heat of Oxidation of Tin, and second paper on the 
Heat of Combination of Acidic Oxides with Sodium 
Oxides ; by W. G. Mixter . . . 229 

XVI.— Neptunite Crystals from San Benito County, Cali- 
fornia ; by W. E. Ford _ ,_ 235 

XVII. — Gravimetric Determination of Silver as the Chro- 

mate ; by F. A. Gooch and R. S. Bosworth 241 

XVIII. — Doppler Effect in Positive Rays; by J.Trowbridge 245 

XIX. — New Armored Saurian from the Niobrara ; by G. R. 

Wieland . . _ 250 

WoLCOTT GlBBS _. 253 

SCIENTIFIC INTELLIGENCE. 

Chemistry and Physics— New Method of Forming Liquid Alloys of Sodium 
and Potassium, Jaubert : Determination of Cerium and other Eare Earths 
in Eocks, Dietrich, 260. — Solubility of Metallic Gold in Hydrochloric 
Acid in the presence of Organic Substances. Awerkiew : The Composi- 
tion of Matter, E. Mulder, 261. — Introduction to the Earer Elements, P. 
E. Browning : Feste Losungen und Isomorphism, G. Bruni : Produc- 
tion of Helium from Uranium, F. Soddy : Charge and Nature of the 
a-Particle, E. Eutherford and H. Geiger : Amount of Water in a Cloud 
formed by Expansion, W. B. Morton, 262. — Permanent Magnetism of 
Copper, J. G. Gray and A. D. Eoss : United States Magnetic Tables and 
Magnetic Charts for 1905, L. A. Bauer, 263. 

Geology and Natural History— Fossilen Insekten und die Phylogenie der 
Eezenten Formen, A. Handlirsch, 263. — Connecticut Geological and 
Natural History Survey, W. N. Eice : Mississippi State Geological Sur- 
vey, A. F. Crider, 264. — Interpretation of Topographic Maps, E. D. 
Salisbury and W. W. Atwood : Zonal Belt Hypothesis, a New Explana- 
tion of the Causes of the Ice Age, J. T. Wheeler : Handbuch der Min- 
eralogie, C. Hintze : Chemische Krystallographie, P. Groth, 265. — Notes 
of a Botanist on the Amazon and Andes, E. Spruce, 266. 

Miscellaneous Scientific Intelligence — Carnegie Institution of Washington. 
Year Book No. 7, 1908, 267.— Eeport of the Librarian of Congress and 
Eeport of the Superintendent of the Library Buildings and Grounds, H. 
Putnam : Harvard College Observatory, E. C. Pickering, 269. — Publica- 
tions of the Allegheny Observatory of the University of Pittsburgh : 
Washburn Observatory of the University of Wisconsin, G. C. Comstock : 
Treatise on Spherical Astronomy, E. Ball : Bulletin of the Mount 
Weather Observatory, W. J. Humphreys, 270, — National Antarctic Expe- 
dition, 1901-1904 : Chemical Constitution of the Proteins, E. H. Aders 
Plimmer, 271. — Standard Algebra, W. J. Milne: Kraft, okonomis the, 
technische und kulturgeschichtliche Studien uber die Machtentfaltung der 
Staaten, E. Eeyer : International Congress of Applied Chemistry : The 
Science Year Book, 272. 

Obituary— TL. G. Seeley, 272. 



VI CONTENTS. 

Number 160. 

Page 

Art. XX. — Permeabilities and the Reluctivities, for very 
Wide Ranges of Excitation, of Normal Specimens of 
Compressed Steel, Bessemer Steel and Norway Iron 
Rods ; by B. O. Peirce _ 273 

XXI. — Ice Movement and Erosion along the Southwestern 

Adirondacks ; by W. J. Miller 289 

XXII. — Estimation of Vanadic and Arsenic Acids and of 
Vanadic and Antimonic Acids, in the Presence of One 
Another ; by G. Edgar 299 

XXIII. — Method for the Iodometric Estimation of Silver 
Based upon the Use of Potassium Chromate as a Pre- 
cipitant ; by F. A. Gooch and R. S. Bosworth 302 

XXIV. — Brown Artesian Waters of Costilla County, Colo., 
their relations to Certain Deposits of Natron or Soda, 
and what they teach ; by W. P. Headden 305 

XXV.— Volumetric Determination of Small Amounts of 

Arsenic ; by L. W. Andrews and H. V. Faer 316 

XXVI. — Preliminaiy Report on the Messina Earthquake of 

December 28, 1908 ; by F. A. Perret . 321 

XXVIT. — New Connecting Link in the Genesis of Fulgur ; 

by C. J. Maury ...... 335 

SCIENTIFIC INTELLIGENCE. 

Chemistry and Physics — Method for Calculating the Boiling-points of 
Metals, F. Krafft and Knocke : Some Properties of the Radium Emana- 
tion, Rutherford, 336. — Method for Preparing Hydrogen Phosphide, 
Matignon and Trannot : The Theory of Valency, J. N. Friend : Recent 
Advances in Organic Chemistry. A. W. Stewart. 337. — Michelson's Ethel- 
Research, E. Kohl : Influence of Pressure upon Thermoelectric Force, H. 
Horig : Aluminum Cell as a Condenser, J. de Modzelewski : Changes in 
the Spectra of Gases submitted to the Magnetic Field, A. Dufour : Die 
elektrischen Eigenschaften und die Bedeutimg des Selens fiir die Elektro- 
technik. C. Ries, 338. — Physics for Secondary Schools, C. F. Adams : 
Elements of Physics. G. A. Hoadley, 339. 

Geology and Natural History — Iowa Geological Survey Report for 1907, S. 
Calvtn : Oklahoma Geological Survey, C. N. Gould, L. L. Hutchison 
and G. Nelson, 339. — Glaciation of Uinta and Wasatch Mountains, W. W. 
Atwood : Glacial Waters in Central New York, H. L. Fairchild : Ground 
Waters of Indio Region, California, W. C. Mendenhall, 340.— Die Alpen im 
Eiszeitalter, A. Penck und E. Bruckner : Geological Survey of Western 
Australia, A. G. Maitland : Die Geologischen Grundlagen der Abstaui- 
mungslehre, G. Steinmann, 341. — Mineralien-Sainmlungen, W. Prendler: 
Jadeite from Upper Burma, A. W. G. Bleeck, 343. — Rubies from Upper 
Burma, A. W. G. Bleeck : New Group of Manganates, L. L. Fermor : 
Die Bliitenflanzen Afrikas, F. Thonner, 344. — Schwendeners Vorlesungen 
iiber mechanische Probleme der Botanik. Holtermann : Plant Study with 
Directions for Laboratory and Field Work, Meier, 345. 

Miscellaneous Scientific Intelligence — Carnegie Foundation for the Advance- 
ment of Teaching ; Report of the President, H. S. Pritchett, and Treas- 
urer, T. M. Carnegie, 346. — Carnegie Institution of Washington, 347. — 
Report of the Superintendent of the Coast aud Geodetic Survey, O. H. 
Tittmanx : Principal Facts of the Earth's Magnetism, 348. 



CONTENTS. Vll 

Niimlber 161. 

Page 

Art. XXVIII. — Weathering and Erosion as Time Measures ; 

by F. Leverett — ._ — 349 

XXIX. — Chalk Formations of Northeast Texas ; by C. H. 

Gordon 369 

XXX. — Devonian of Central Missouri ; by D. K. Greger.. 374 

XXXI. — Volumetric and Gravimetric Estimation of Thal- 
lium in Alkaline Solution by Means of Potassium 
Ferricyanide ; by P. E. Browning and H. E. Palmer. _ 379 

XXXII.— Descriptions of Tertiary Insects, VI; by T. D. A. 

Cockerele ... . 381 

XXXIII. — Divided Lakes in Western Minnesota ; by R. F. 

Griggs ---;---. .- - : 3 8 8 

XXXIV. — Heat of Formation of Titanium Dioxide, and 

third paper on the Heat of Combination of Acidic Oxides 

with Sodium Oxide ; by W. G. Mixter 393 

XXXV.— Note on Crystal Form of Benitoite ; by C. Palache 398 
XXXVI. — Alamosite, a new Lead Silicate from Mexico ; by 

C. Palache and H. E. Merwin 399 

.XXXVII. — Absence of Polarization in Artificial Fogs ; by 

C. Barus ... .- 402 

SCIENTIFIC INTELLIGENCE. 

Chemistry and Physics — Prussian Blue and Turnbull's Blue, Muller and 
Stanisch : Chemistry of the Eadio-active Elements, Stromholm and 
Svedberg, 403. — Hydrogen Silicides, Lebeau : Determination of Boron, 
Copatjx and Boiteau, 404. — Canal Rays, J. Stark and W. Steubing : 
Zeeman Effect of Mercury Lines, P. Gmelin : Presence of Eays of High 
Penetrability in the Atmosphere, T. Wulf : Use of Radiometer for Observ- 
ing Small Pressures, J. Dessar, 405. — Wireless Telegraphy and Telephony 
Popularly Explained. W. W. Massie and C. R. Underbill : Introduction 
to the Science of Radio-activity, C. W. Raffety, 406. 

Geology and Natural History — Publications of the U. S. Geological Survey, 
G. O. Smith, 406. — Economic Geology of the Georgetown Quadrangle, 
Colorado, J. E. Spurr and G. H. Garrey, 408. — Geology of the Gold 
Fields of British Guiana, J. B. Harrison, 409. — Eruption of Vesuvius in 
April, 1906, H. J. Johnston-La vis : Essai sur la Constitution g^ologique 
de la Cuyane hollandaise (district occidental), H. Van Capelle : Text- 
Book of Petrology, F. H. Hatch, 410. — Classification of the Plutonic 
Rocks, F. H. Hatch, 411. — Elements of Optical Mineralogy, Winchell : 
Geology of the Taylorsville Region, California, J. S. Diller, 412. — Explo- 
rations in Turkestan : Glacial Bowlders in the Blaini Formation, India, T. 
H. Holland : Guadalupian Fauna, G. H. Girty, 413. — Cambrian Sections 
of the Cordilleran Area, C. D. Walcott : Mount Stephen Rocks and Fos- 
sils, C. D. Walcott : Devonian Fishes of Iowa, C. R. Eastman, 414. — 
Unterkiefer des Homo Heidelbergensis aus den Sanden von Mauer bei 
Heidelberg, O. Schoetensack, 415. — Commercial Products of India, G. 
Watt, 417.— Forest Flora of New South Wales, J. H. Maiden, 418. 

Miscellaneous Scientific Intelligence — National Academy of Sciences, 418. — 
Allgemeine Physiologie, M. Verworn : Man in the Light of Evolution, 
J. M. Tyler, 419. — Harvard College Observatory, E. C. Pickering : Pub- 
lications of the Allegheny Observatory of the University of Pittsburgh : 
Brooklyn Institute of Arts and Sciences, 420. 

Obituary — Dr. Persifor Frazer, 420. 



Vlll CONTENTS. 



Number 162. 

Page 

Art. XXXVIII. — Quartz as a Geologic Thermometer ; by 

F. E. Weight and E. S. Laesen _ 421 

XXXIX. — Precipitation of Copper Oxalate in Analysis ; by 

F. A. Gooch and H. L. Ward _ _ „ . 448 

XL. — Yakutat Coastal Plain of Alaska ; A combined ter- 
restrial and marine Formation ; by E. Blackwelder _ 459 

XLI. — Pyrite Crystals from Bingham, Utah ; by A. F. 

Rogers ._ 467 

XLII. — Composition of Stony Meteorites compared with 
that of Terrestrial Igneous Rocks, and considered with 
reference to their efficacy in World-Making ; by G. P. 
Meeeill _. 469 

XLIIL— Notes on the Family Pyramidellidae ; by K. J. 

Bush, Ph.D. 475 

SCIENTIFIC INTELLIGENCE. 

Chemistry and Physics — Liquid and Solid Radium Emanation, Gray and 
Eamsay : Gases Evolved by the Action of Cupric Chloride upon Steels, 
Goutal, 485. — Radiation of Potassium Salts, E. Henriot : Course of 
Qualitative Chemical Analysis of Inorganic Substances, 0. F. Tower, 
486. —Suggested Method of Ascertaining the Existence of Chlorophyll on 
Planets, N. Umow : Condensation of the Radium Emanation, Ruther- 
ford : Electric Origin of Molecular Attraction, W. Sutherland, 487. — 
Physical Measurements, A. W. Duff and A. W. Ewell, 488. 

Geology and Natural History — Monograph on the Higher Crustacea of the 
Carboniferous Rocks of Scotland, B. N. Peach, 488. — Fossils from the 
Silurian Formations of Tennessee, Indiana and Illinois, A. F. Foerste : 
Illinois State Geological Survey, C. W. Rolfe, R. C. Purdy, A. N. Tal- 
bot and I. 0. Baker : Wisconsin Geological and Natural History Survey, 
E. A. Birge and W. 0. Hotchkiss : A new Locality of Diamonds in Africa, 
E. Kaiser, 489. — Interferenzerscheinungen im polarisirten Licht, H. 
Hauswaldt : Complete Mineral Catalog, W. M. Foote : Introduction to 
the Study of Rucks, L. Fletcher : Determination of Rock-forming Min- 
erals, A. Johannsen, 490. — Trees, A Handbook of Forest Botany for the 
Woodlands and the Laboratory, H. M. Ward : Mendel's Principles of 
Heredity, W. Bateson, 491. — Catalogue of the Lepidoptera Phalaenae in 
the British Museum, Vol. VII ; Catalogue of the Noctuidse, G. F. Hamp- 
son, 492. 

Miscellaneous Scientific Intelligence — Bulletin of the Mount Weather Obser- 
vatory, 492. — Field Columbian Museum, Chicago : Geographical Tables, 
Albrecht : An Astronomer's Wife, A. Hall : Anciennes Meures, Mas- 
cart, 493. 
Index, 494. 



piT Gyrus Auier, 

Librarian U. S. Nat. Museum. 



VOL. XXVII. JANUARY, 1909. 



Established by BENJAMIN SILLIMAN in 1818. 



THE 

AMERICAN 

JOURNAL OF SCIENCE. 

Editor: EDWAED S. DANA. 

ASSOCIATE EDITORS 

Professors GEOKGE L. GOODALE, JOHN TKOWBKIDGE, 
W. G. FAKLOW and WI. M. DAVIS, of Cambridge, 

Professors ADDISON E. VEKKILL, HOKACE L. WELLS, 
L. V. PIKSSON and H. E. GKEGOEY, of New Haven, 

Professor GEOKGE F. BAEKEK, of Philadelphia, 
Professor HENEY S. WILLIAMS, of Ithaca, 
Professor JOSEPH S. AMES, of Baltimore, 
Mr. J. S. DILLEE, of Washington. 



FOURTH SERIES 

No. 157— JANUARY, 1909. 
VOL. XXVII— [WHOLE NUMBER, CLXXVIL] 

WITH PLATE I. 



NEW HAVEN, CONNECTICUT. 

1909. 

THE TUTTLE, MOREHOUSE & TAYLOR CO., PRIXTERS, 123 TEMPLE STREET. 



Published monthly. Six dollars per year, in advance. $6.40 to countries in the 
Postal Union ; $6.25 to Canada. Remittances should be made either by monr-y orders, 
registered letters, or bank checks (preferably on New York banks). / ^'2,^ 



%., 



RARE CINNABARS FROM CHINA. 

We have still left a small lot of these remarkable Cinnabar specimens 
which were described and illustrated in the November issue of this Journal. 
Write for prices and illustrated pamphlet. 

NEW ARRIVALS. 

Euclase, Capo do Lane, Brazil ; Terlinguite, Terlingua, Texas ; Eglestonite, 
Montroydite, Terlingua, Texas ; Patronite, S. A. ; Uraninite, crystal in 
matrix, Portland, Conn.; Benitoite, San Benito Co., Cal.; Neptunite, Cal.; 
Crookesite, Hedenbergite, Sweden ; Lievrite, Elba ; Polybasite, Hungary and 
Durango, Mexico ; Josephinite, Oregon ; Herderite, Maine ; Smithsonite, 
Kelley, N. M. ; Calif ornite, Tulare Co. , Cal. ; Cobaltite, loose crystals, and in 
matrix. Cobalt, Ont., and Tunaberg, Sweden; Apatite, Auburn, Maine; 
Yivianite, large crystals, Colo. ; Vanadinite, Kelley, N. M. ; Monazite, Portland, 
Conn.; Olivenite, Utah; Sartorite, Canton Wallis ; Jordanite, Binnenthal ; 
Mohawkite, Algodonite, Domeykite, Michigan ; Crocoite, Siberia and Tas- 
mania ; Cinnabar, Cal. and Hungary ; Dioptase, Siberia ; Diopside, with 
Essonite, Ala, Piedmont ; Embolite, Australia ; Gypsum, twin crystals, 
Eisleben, Thuringia ; Diamond, in matrix, New Vaal Eiver Mine, South 
Africa ; Brochantite on Chrysocolla, Utah ; Pink Beryl, small and large, 
Mesa Grand, Cal.; Kunzite, small and large, Pala, Cal.; Sphene, Binnenthal, 
Titanite, Tilly Foster, N. Y. ; Tetrahedrite, Utah and. Hungary; Eealgar, 
Hungary ; Opal, Caribou Paver, Queensland ; Octahedrite, var. Wiserine, 
Binnenthal ; Heulandite, Iceland ; Paper Calcite, Saxony ; Torbernite, Eng. ; 
Bismuth, native. Cobalt, Ont. and Conn. ; Silver, native with Breithauptite 
and Smaltite, polished, Ont.; Emerald, loose crystals and in matrix, Ural 
and Bogota ; Topaz crystals with rare planes, Ural ; Zircon crystals, loose, 
Ural ; Green and Cinnamon Garnets, Minnot, Maine ; Vesuvianite, Poland, 
Maine, Italy and Tyrol ; Zeolites, beautiful specimens from Erie Tunnel, 
Patterson and Great Notch. 



CUT GEMS. 

We have left over from our Christmas trade a very fine and varied lot of 
large and small Cut Gems, which we are offering at unheard of low prices. 
We name a few below : — 

Garnets, green and red ; Aquamarines ; Zircons, all shades ; Sapphires, 
all shades ; Star Sapphires and Star Eubies ; Chrysoberyl, Cats-eye ; Spinels, 
all shades ; Topaz, pink, blue, brownish and golden color ; Pink Beryl ; 
Sphene; Tourmaline, all shades ; Amethyst, Siberia, royal purple color ; 
Andalusite ; Star Quartz ; Peridote ; Opal matrix, Mexico and Australia ; 
Precious Opal, Australia, Mexico and Hungary ; Hyacinth ; Turquoise, Mex- 
ico and Persia : Kunzite ; Reconstructed Rubies and Sapphires ; Emeralds ; 
Opal Carvings, such as pansies, vine leaves with bunches of grapes, and 
other small Opal novelties ; Antique and Modern Cameos ; Antique, Mosaic 
and other semi-precious stones. 

Let us know your wants, and we will send the specimens on approval to 
you. 

A. H. PETEREIT, 

81—83 Fulton Street, New York City. 



THE 



AMERICANJOURNALOFSCIENCE 

[FOURTH SERIES.] 



Art. I. — Diopside arid its Relations to Calcium and Magne- 
sium Metasilicates ; by E. T. Allen and W. P. White. 
With Optical Study : by Feed. Eugene Wright and 
Esper S. Larsen. (With Plate I.) 

The pyroxenes were chosen some time since as a good 
subject for laboratory investigation, both by reason of their 
geologic importance and becanse their comparative stability 
and simplicity of composition seemed to offer relatively little 
difficulty in their synthesis and study. Diopside is the sim- 
plest of this group of minerals, but before even this could be 
studied satisfactorily, a detailed investigation of both calcium 
and magnesium metasilicates was found necessary, a full 
account of which has been given elsewhere.* Here it will be 
sufficient to state that calcium silicate exists in two crystal 
forms, one of which, the mineral wollastonite (/3-form), is stable 
up to about 1190°, when it reverts to a pseudo-hexagonal form 
(a-form) which melts at 1512°. The case of magnesium metasil- 
icate is more complex. There is one form (monoclinic) strongly 
resembling the pyroxenes, both optically and crystallographic- 
ally, which is stable up to 1365° (/3-form). Here it passes 
over into an orthorhombic form (a-form) recalling forsterite 
(Mg 2 Si0 4 ) in its habit and optical properties. It melts at 
1524°. Three other forms exist, viz., the minerals enstatite, 
kupff'erite, and a monoclinic amphibole similar to the latter. 
All are monotropic, and change into the /3-form when heated 
to a sufficiently high temperature. 

The question of the relation of diopside to its component 
salts, calcium and magnesium silicate, we undertook to settle 
by determinations of the melting point and specific volume 

* Allen, White and Wright, this Journal, xxi, 89, 1906; Allen, Wright 
and Clement, this Journal, xxii, 385, 1906. Their form, then undiscovered, 
is described for the first time in these pages. 

Am. Jour. Sci.— Fourth Series, Vol. XXVII, No. 157.— January, 1909. 
1 



2 Allen, etc. — D topside and its Relations 

curves, relying on very careful microscopic examinations to 
make sure of the solid phases which separated from the molten 
solutions. Since we are concerned here with equilibrium 
conditions, it will be evident that the monotropic forms of 
magnesium silicate do not enter into the problem.* In addi- 
tion to this specific problem, viz., the relations existing between 
the calcium and magnesium silicates, we planned to study the 
most important properties of diopside and its transformations 
on heating, if any. 

Many of the methods used in this investigation have been 
already described in previous papers from this laboratory ;f 
others have been devised in the course of the work. 

Preparation of the Mixtures. — Quartz, magnesia and cal- 
cium carbonate were mixed in the proper proportions and 
melted in large covered platinum crucibles. The molten 
charges were chilled by dipping the crucibles into water, 
cracked out of the crucibles, crushed fine, and remelted to 
ensure thorough mixing. All the materials were carefully 
examined for impurities, and we are able to state with cer- 
tainty that the products contained no more than Ol per cent to 
0'2 per cent of foreign matter, except in certain cases where 
the same preparation was repeatedly melted and crushed, when 
as much as 0*3 per cent impurity was sometimes found. 

We had considerable trouble in getting magnesia sufficiently 
pure for our purposes. The chief impurity is generally lime, 
of which all the samples examined by us contained 0'5 per cent 
or more. At our request, the firm of Baker & Adamson under- 
took the problem of preparing pure magnesia in quantity, and 
succeeded in making a " basic carbonate" in which the mini- 
mum quantity of lime amounted to about "07 per cent, or 015 
per cent of the calcined oxide. A sample of magnesium 
metasilicate prepared from one lot of this carbonate gave by 
direct test '12 per cent lime. In mixtures very high in mag- 
nesia a correction was generally made for this impurity, i. e., 
the proper deduction was made from the quantity of lime 
required by the particular mixture in hand. For the sake of 
control we analyzed two of the preparations and the results 
are given below. 

At one time during the investigation, it was suspected that 
some loss might be caused by the strong blast of the gas fur- 
nace in which the constituent powders were first melted together 
in spite of the fact that the crucible was carefully covered. 
The magnesia being by far the lightest of the constituents 

* Viktor Poschl incorrectly designates the series " ensftxfo'te-diopside," 
Centr. Min., 1906, p. 572. 

fThis Journal, xix, 125, 1905; xxi, 89, 1905; Tschermak ; s Mitth., xxvi, 
169, 1907. 



to Calcium and Magnesium Metasilicates. 3 

(magnesia, calcium carbonate, quartz), its loss would be dispro- 
portionate to its weight in the mixture. But the analysis of 
the 60 per cent MgSi0 3 mixture, in which the magnesia was 
slightly high, shows that the suspicion was unfounded. 

Found Cal. 

1. 10#MgSiO 3 ._ Si0 2 52-75 52-67 

90$ CaSi0 3 Caa ._ 43-52 43-33 

MgO--- -- 4-06 4-00 

Fe 2 3 etc '11 

100-44 100-00 

2. 60$MgSiO 3 Si0 2 56*80 56'73 

40$ CaSi0 3 CaO 19-12 19-26 

MgO 24-11 24-01 

Fe 2 3 etc. '13 



100-16 100-00 

Thermal Study. — The thermal study was carried on, as in 
other published work from this laboratory, by Frankenheim's 
method, melting point curves being used exclusively. Freez- 
ing-point curves are unreliable on account of the undercooling, 
which, even in substances which crystallize as readily as the 
metasilicates of calcium and magnesium, is often considerable. 
The mixture under investigation was heated in a platinum 
crucible, in an electric resistance furnace fed from a storage 
battery, the temperatures were read to tenths of a degree by a 
Le Chatelier thermoelement and a potentiometer, and evalu- 
ated in terms of a standard saturated cadmium cell. 

Since the work here described continued over a considerable 
period of time, the methods used varied somewhat as succes- 
sive improvements were introduced. In the earlier portion of 
it, charges of about 25 grams were used, and the thermoelement 
was protected by a porcelain tube with a platinum jacket. A 
control element was also employed to detect any irregularity in 
the heat supply.* Soon after the work was begun, the practice 
was also instituted of comparing the thermoelements with 
standards in order to correct for their deterioration, f due to 
contamination with iridium from the platinum crucibles and 
furnace coil. This has now been greatly reduced by using 
specially pure platinum within the furnace, but occasional com- 
parisons with standard elements are still necessary 4 Fre- 
quently the thermoelements were inserted together in the 
porcelain tube and used to read the same melting point. The 

*This Journal, xxi, 94, 1906. 

•{•The Constancy of Thermoelements, W. P. White. Phys. Zeitschr., viii, 
325, 1907. 
%W. P. White, Phys. Rev., xxv, 336, 1907. 



4: Allen, etc. — Diopside and its Relations 

readings when separately corrected then seldom varied -J- , 
which was about thq accuracj^ with which any two working ele- 
ments could be compared. The variations in the results, some- 
times amounting to 2°, even with the- more sharply melting 
substances (the end members, the eutectics, and diopside), were 
therefore due to uncertainty in locating the melting point, and 
not to the temperature measurements. 

In the latter part of the work, charges of only 2 J grams 
were used, in very small crucibles of pure platinum, in which 
the naked thermoelements could safely be immersed. Under 
these conditions and with proper precautions, the variations 
rarely exceeded \° . The small crucibles could also be more 
easily moved in the hot furnace, chilled in cold water, etc., and 
a wider range in the conditions of crystallization obtained. 
Furthermore, the thermal lag of the diminutive charges was 
very small, and complex thermal phenomena could be analyzed 
or separated with much greater certainty than with larger cru- 
cibles. The small crucibles were mainly used in investigating 
those portions of the field in which the interpretation of the 
phenomena had proved particularly difficult. They were used, 
however, in a sufficient number of cases to afford a thorough 
control of the earlier determinations made with 25 grams, and 
showed that these were accurate within their own accidental 
errors. 

The agreement obtained between observations of the same 
point in both the earlier and later work is well illustrated in 
the following table : 

Table I. 

Agreement of Observations in Earlier and Later Work. 

Temperatures in microvolts. About 13 microvolts to 1°. 

70#MgSiO 3 : 30#CaSiO 3 . 28$ MgSi0 3 ; ?2$CaSi0 3 . 

25 gr. charge. 2£ gr. charge. 

June, 1905 14,147 Oct. 1907. 13,772 

130 776 

110 777 

138 779 

148 

127 



Mean 14,133 13,776. 

Average deviation, 11 mv. 2 mv. 

Extreme " 37 mv. 7 mv. 

These particular instances were selected for illustration as 
being, each in its class, the one showing the largest number of 
strictly comparable observations. Each one is typical, and 






to Calcium and Magnesium Metasilicates. 5 

neither is the best that could be given. Each represents three 
different charges. 

A word as to the exact significance of the accuracy here indi- 
cated may avoid some confusion. 

In comparative measurements with very small crucibles (2*5 
grams) and bare thermoelements, the relative accuracy obtain- 
able by the same observer with the same apparatus is J° or 
better ; with the larger crucibles (25 grams) and jacketed ele- 
ments, differences may reach 3°. This is shown in Table I, 
page 1. 

The absolute value of the measurements is not established 
with this accuracy. Different observers, indeed, working inde- 
pendently with different apparatus, ought not to differ much 
in their results through differences in experimental methods, 
provided these are not positively faulty ; that is, provided the 
thermal junction is actually at the temperature of the melting 
material. This is shown by the agreement of our earlier and 
later results, obtained with highly differing methods, and also, 
perhaps more conclusively, by the agreement found between 
Dr. Day's special form of element* and the ordinary jacketed 
element. Differences between independent observers are, how- 
ever, to be expected in the thermoelement calibration, partic- 
ularly at the present time, when a long extrapolation is 
necessary. Practically all temperature measurements above 
the melting point of copper (1084° C, Reichsanstalt Scale) are 
now obtained by extrapolating some simple function which 
has been experimentally established for temperatures below 
that point. With the same function and different elements, 
temperatures usually do not vary more than 2° C. at 1,500°. 
With different functions, the extrapolated curves may diverge 
as much as 30° C. at 1,500°. An error of 0'5° in the determi- 
nation of the copper point itself may cause an error of 4° in 
the extrapolated curve at 1,500°. All these are differences in 
the interpretation of the experimental measurements, which 
will probably continue to cause considerable differences 
between the results of different observers until the gas ther- 
mometer scale is extended to that temperature with sufficient 
accuracy, after which they can be readily recomputed to the 
established scale. The temperatures here given are com- 
puted from the Reichsanstalt Scale. In brief then, the accuracy 

* Carnegie Institution Publication No. 31, p. 25. Our value for magne- 
sium silicate is 3° higher than that found by Allen, Wright and Clement, 
but this difference is only a little greater than the accidental variations 
occurring in the present work. With charges of the same size, errors in the 
method seem most likely to occur, if at all, as the result of a very uneven 
temperature distribution in the furnace or of insufficient immersion of the 
thermoelement in the charge. In a few cases where the platinum jacket 
dropped down so as to rest against the bottom of the crucible, which was 
cooled by contact with the pedestal below it, the melting point came about 
5° too low. 



6 Allen, etc. — Diopside and its Relations 

of the determinations is shown by comparing the measure- 
ments among themselves.' Their interpretation in terms of 
an absolute scale depends upon an agreement among observers, 
and will vary from time to time whenever more accurate funda- 
mental observations are available. At the present time the 
Reichsanstalt Scale is generally accepted. 

The melting points measured, like those obtained in this lab- 
oratory with the feldspars, and by workers in high tempera- 
tures generally, are not entirely sharp, even with substances 
which theoretically should melt at a strictly constant temper- 
ature. Indications that melting has already begun invariably 
appear on the thermal curve 20° or 30° below the melting point 
proper, and the region of strongest absorption of heat is distri- 
buted over an interval of from 2° to 3°. The main cause of 
this phenomenon in the present case probably lies neither 
in any unusual molecular viscosity attending the change 
of state* nor in experimental error of the temperature obser- 
vations. It may be due to the slight impurity ('1 to *2 per 
cent) which chemical analysis shows to be present, even in the 
most carefully prepared artificial mixtures. Similar curves are 
obtained with ice to which 2 per cent or 3 per cent of salt has 
been added, and at an absolute temperature five or six times 
as high the same effect should be produced by about one-thirtieth 
as much impurity. f The highest portion, that is, the end, 
of the melting interval is taken as the melting point, since this 
is probably the nearest attainable approach to what would be 
the sharp melting point of the substance unaffected by traces 
of impurity. 

The Melting Point Curve from jper cent— lfi'3 -per cent 
MgSi0 2 . — The temperature-time curves were taken on a series 
of mixtures of the metasilicates 5 per cent to 10 per cent apart 
except in critical parts of the curve, where shorter intervals were 
chosen. The temperatures at which the heat absorptions took 
place are plotted in fig. 1. The results in the case of about 
one-third of the mixtures were controlled by repeating the 
observations on several different preparations, and since the 
thermal phenomena in certain parts of the curve w T ere rather 
complicated, a considerable number of the observations were 
many times repeated, making in all nearly four hundred inde- 
pendent determinations. 

By reference to the diagram (fig. 1) it will be seen that the 
addition of magnesium silicate to pure calcium silicate (pseudo- 
wollastonite) lowers the melting point rapidly, the curve fol- 

* This, of course, is not true of the feldspars, quartz, etc. 
f In accordance with the Van't Hoff-Eaoult formula, where A, the 
depression of the freezing point due to impurity, equals '02 T- 



to Calcium and Magnesium Metasilicates. 7 

lowing a slightly curved line to the eutectic point, 1348°, at about 
28 per cent magnesium silicate. The curve then rises much 
more gradually to a maximum of 1380° at the composition of 
diopside, CaSi0 3 .MgSiO s , which contains 46*3 per cent mag- 
nesium silicate. Let us first consider so much of the curve by 
itself. The solid phases, which separated when the mixtures 
from per cent— 46*3 per cent magnesium silicate were crys- 
tallized as nearly as possible under equilibrium conditions, were 
proved by microscopic analysis to be only diopside and pseudo- 
wollastonite, making of course the proper allowance for a small 
mutual solubility. The latter amounted, as we shall see, to 
about 3 per cent diopside in the pseudo-wollastonite, and less 
than 3 per cent of the latter in diopside. 

The end members, pseudo-wollastonite and diopside, and also 
the eutectic mixture, melt at a single temperature, with the 
reservation just made ; that is, as sharply as we have yet ob- 
served with any silicate. The other mixtures show two phe- 
nomena : (1) the melting of whatever amount of eutectic is 
present in the mixture ; and (2) as the temperature rises, the 
gradual solution in the melted eutectic of the excess component 
(diopside or pseudo-wollastonite, as the case may be), which lasts 
till the melting point curve ABC is reached. The temperature- 
time curves clearly show continuous change in character from 
the end members to the eutectic composition. Near the end 
member the predominant phenomenon is the upper point, 
which resembles the sharp melting of the pure component 
itself ; the eutectic melting is of course small. It is clearly 
distinguishable, however, even when the mixture contains less 
than 1 per cent of eutectic. In mixtures farther from the end 
members in composition, the absorption of heat immediately 
above the eutectic melting is perceptible, and the highest point 
gradually takes on more and more the character of the end of this 
absorption and less and less that of a separate and independent 
melting point. Although, strictly speaking, this upper melting 
is always a solution, the temperature at which it terminates did 
not show, under the conditions described, any change with the 
rate of heating. In order to test this question, the rate was in 
several cases altered from 1° to 3° per minute, but the result- 
ing effect on the upper point was less than the accidental errors 
of the separate determinations. As the eutectic composition 
is approached and the phenomenon has still more the character 
of a solution and less that of a melting, the upper point 
becomes less and less distinct. 

Determination of the Eutectic Composition. — In the imme- 
diate neighborhood of the eutectic composition the difficulty of 
determining the upper point is still more increased by another 
phenomenon. The rapid temperature rise which immediately 



Allen, etc. — Diopsicle and its Relations 

Table II. 

Temperature- Concentration Curve. 



Weighty 
MgSiOs 


Upper 
Point 





1510° 

1511 

1510 


1 





2 





5 

8 


1489 
1481 
1483 


10 


1468 
1466 
1466 


20 


1405 
1397 
1406 
1407 


24 





26 





28 





30 





31 






32 



36 



Eutec- 
tic 



1343 c 

1345 

1343 

1346 
1343 
1347 
1346 

1350 
1350 

1348 
1348 

1348 
1348 
1348 

1348 
1349 
1349 

1349 
1350 
1350 

1348 
1349 
1350 

1348 

1348 



Inver- 
sion 



1190 c 

1245 
1273 

1327 



Weighty 
MgSiOs 

40 


Upper 
Point 

1374° 


Eutec- 
tic 


1347° 


45 


1-380° 


1343 


46-3 


1380 
1381 
1379 




47-5 


1381 
1381 


-- 


500 


1381 
1380 


\~~. \ 


55 


1378 


.... 


60 


1378 
1377 


— 


65 


1377 

1378 


— 


70 _ 





1376 


72 


1409 


1375 


73 


1408 
1412 


1373 


75 " 


1416 


1372 


80 


1452 
1453 
1452 


1373 
1375 


90 


1495 
1484 

1487 


1375 


95 


1512 


1387 


98 








100 


1524 
1524 





— 









Inver- 
sion 



1363 



1364 
1367 

1367 

1362 
1360 



to Calcium and Magnesium Metasilicates. 9 

Fig. 1. 

















































A 






































y/i 


E 


1510° 


t\ 










































1480° 


; « 


\ 








































I ' 


\ 






























f 








1450° 




\ 












































\ 


V 




































1420° 








\ 






















/ 




















\ 






















/ 


i 




8 








1390° 










\ 




















/ 


















i 




\ 










C 








yPl, 












v\ 














\ 


\ 












7 




•r 


o 


i 


I 9 






xy 


W 


~ 1360° 


<U c 


!> « 


3 














xH 








o 


V 


/ 


2 


I 














































S 1330° 


; 1 








B 






i 


























I /* 








































1300° 




i 




















6 
































3 
































4270° 


\ 
































/0 




















































1240° 






















































































1240° 






















































































4480° 


K^ \ 


— - 


.__ 


--- 


■-- 


-- 




. .11 




























5 V 


i 






4 


































44-50° 


L 
















Q 


R 








Is 












y\ 


z 



40 
CaSiOs 



60 



Concentration'. 
Principal Areas of the Diagram. 

AHB a-CaSi0 3 + liquid. 

B1C Diopside + liquid. 

HIMO a-CaSiOs + diopside. 

OPQM ,3-CaSi0 3 + diopside. 

KOPL Mix-crystals diopside in ,3-CaSi0 3 . 

CTUSR Mix-crystals of MgSi0 3 in diopside. 

CDT Mix-crystals of varying composition + liquid. 

DE\T a-MgSiOa + liquid. 

TUXV a-MgSi0 3 + diopside mix-crystals. 

UXYS ,3-MgSiC) 3 + diopside mix-crystals. 



90 100 

MgSiOs. 



10 



Allen, etc. — Diopside and its delations 



follows the principal melting now extends beyond the upper 
point and completely overwhelms it. The difficulty of deter- 
mining a small residual melting is in striking contrast to the 
ease with which a slight absorption of heat can be detected 
below the principal melting.* Owing to this masking of the 
upper point, a direct determination of the eutectic composition 
is impossible by the ordinary method; that is, we cannot dis- 
tinguish within several per cent the mixture which melts leav- 

Fig. 2. 



1400° 


























































































































/ 






/ 
































































/ 


/ 


1 




/ 






























































7 


4- 




/ 






i 


A 


- 








A 














































\ 




J 












































< 


Y 








V 


t 


















i 




































/ 


? 












> 


/ 


















H 


V 


' 
































































/ 


S 


f 


































































i 


? 


Y 












1350° 
















[fi 












r— < 


1 1 


h 
















p-t 




e=S 


►— 


" i J 


7 






























r — 1 






















































1340° 


/ 


/ 


/ 


> 




€ 


















/ 


/ 


i 






> 


s. 


s 
































i 






/ 


















<< 


h 


/ 








V 


' 


































1330° 




























( 






' 


A 


i 















































































































Temperature-time curves of mixtures in the vicinity of the eutectic com- 
position (pseudo-wollastonite-diopside). 

I 24$ MgSi0 8 . II. 26% MgSi0 3 . III. 28$ MgSi0 3 . IV. 32% MgSi0 3 . 
v . 36% MgSi0 3 . 

ing no excess component, since a number of mixtures appear 
to do so. This is well shown by the temperature-time curves 
in fig. 2, which were made with especial care with reference 
to this very point. The upper melting can hardly be distin- 
guished in the 24 per cent mixture, though this contains over 
8 per cent of the component in excess. ' (The point is in fact 
so faint that we could hardly distinguish it from the minute 
irregularities in the temperature curve of the furnace were it 
not for the fact that it lies on the curve ABC, fig. 1, which is 
well defined in other mixtures farther removed from the 
eutectic.) For locating the eutectic two methods still remain. 

* It was undoubtedly the difficulty of detecting the upper points in this 
region and the failure to appreciate how easily such an experimental diffi- 
culty might arise which led R. Freis (Neues Jahrb. Min., Beil. Bd. xxiii, 76, 
1907) to describe a similar series of melting points as forming a curve with 
the two inclined portions separated by a horizontal branch of considerable 
extent. 



to Calcium and Magnesium Metasilicates. 



11 



The first is to extrapolate downward the liquidus curves AB 
and BC in fig. 1, and take the mean of the points where they 
intersect the eutectic line HI. The accuracy of this extrapo- 
lation is limited by the uncertainty of the liquidus curve for 
some distance on each side of the eutectic composition. We 
found with these particular silicates that the results agreed to 

Fig. 3. 



1379° 



1377° 



1375° 



1373° 



1371° 



1369° 



1367° 

























/ 


^ 






































/ 


V 






































/, 


Y 


































e 




/ 


V 






































/ 


7 


v 


- 


































h > 


f 




































$ 


in 


// 


O? 


) 
























i 










$ 


// 


C 


f 


































l V 


/? 


M 






































A 


7 


V 


































X 


V > 


V 


, 




































A 


ft 






































^ 


'/ 









































1361° 


































/ 


s 










































/ 


^ 




/ 








1359° 


























\ 


■ / 


V 






) 


































s ' 


V 


/ 


















1357° 




















•V 


^b 


•/■ 








































% 


V 1 


/ 






^ 


/ 
















1355° 














>s 




/ 


/ 










































/ 


i 










7 


















1353° 














/ 










;.y 


w 
































/ 




<S 






























1351 c 










/ 






> 


/ 


































/ 


/ 




s 


S 
































1349° ' 















































Time : \ minute. 

Location of the eutectic composition. 

Curves I-I and Il-II, 28 per cent and 30 per cent MgSi0 3 just above the 
eutectic point. 
Curves I'-I' and II'-II' the same at higher temperature. 

5 per cent with a probable error of not over 2 per cent in the 
eutectic composition. 

A more accurate determination seemed for some time to in- 
volve considerable difficulty, but a satisfactory solution of the 



12 Allen, etc.— Diopside and its Relations 

problem was finally obtained in the following way: Two small 
crucibles containing charges slightly different in composition 
were pnt together in the furnace and brought to such a temper- 
ature that the main eutectic melting was completed but the 
excess component, if any, still left undissolved. Then by sud- 
denly increasing the furnace current and making alternate 
temperature readings of the two crucibles at intervals of 15 
seconds, the relative heat absorptions in the region immediately 
above the eutectic melting point could be determined, and thus 
in a very few trials of different mixtures the eutectic composi- 
tion could be located. The results of such a determination are 
shown in curves I-I and II-II, fig. 3. The temperature in 
II rises more slowly than in I, because of an absorption of 
heat due to the melting of the excess component. The curves 
were continued to a higher temperature to make sure that this 
difference was really due to this cause and not to the relative 
position of the two crucibles in the furnace or to some other 
accidental circumstance. The curves V and IF are now paral- 
lel, confirming the conclusion that their previous divergence 
was due to a melting in II which is now over. It will be seen 
from this result that the 28 per cent mixture lies nearer the 
eutectic than the 30 per cent. A similar comparison of the 
28 with the 26 per cent also showed that the eutectic lies 
nearer the former. These experiments are sufficient to locate 
the point well within 1 per cent at the 28 per cent mixture. 
The previous work had indicated that the 28 per cent was 
probably the eutectic composition. Calculated in terms of 
diopside, the eutectic mixture would contain in round num- 
bers 60 per cent diopside : 40 per cent pseudo-wollastonite. 

Diojpside. — As previously stated, the melting-point of diop- 
side is a maximum on the melting point curve, which fact, 
taken in connection with the microscopic homogeneity at this 
composition and the occurrence of a eutectic point on either 
side of it, proves that it is a compound in stable equilibrium 
with its own liquid. Abundant confirmation of this conclu- 
sion is found in the form of the specific-volume curve, p. 27, as 
well as in the close agreement of the composition of natural 
diopside with the rational formula CaSi0 3 .MgSi0 3 . As an 
example, we give an analysis of an exceptionally pure specimen 
from a metamorphosed limestone which occurs in Ham Island, 
Alaska. The composition of the anhydrous substance is also 
given for the sake of fairer comparison. This is justifiable, 
for the water is not chemically combined, a conclusion thor- 
oughly established by the fact that the substance remains 
homogeneous as the water escapes. 



to Calcium, and Magnesium Metasilicates. 13 

Diopside from 

Ham Island, 

Alaska 

Si0 2 54-65 

Fe.O, -13 

CaO 25-27 

MgO 18-78 

Na o _ -03 

K o -07 

HO 1-45 



Cal. to the 


Cal. for 


anhydrous 


CaSiO-3 


condition 


MgSiOs 


55-46 


53-63 


•13 


. _ _ , 


25-64 


25-82 


19-06 


18-55 


•03 





•07 












100-38 100-39 100-00 

When diopside is prepared by melting together lime, mag- 
nesia and silica in the proper proportions, it crystallizes readily, 
bat not so rapidly as its constituents calcium and magnesium 
silicates, forming a dense white homogeneous mass with con- 
spicuous cleavage. The optical constants, a full account of 
which is given in the second part of this paper, leave no room 
for doubt of the identity of the substance. 

Measurable crystals was prepared by crystallizing this prod- 
uct from molten calcium chloride-Le Chartier's method.* This 
may be very conveniently done in platinum as described in a 
previous paper on magnesium silicate. f It is not possible, 
however, to perform the operation in hydrochloric acid gas, as 
was there done, for the diopside is then decomposed into cal- 
cium chloride, tridymite and magnesium pyroxene. The last 
two products were identified optically without difficulty. The 
calcium chloride should first be melted in the sealed crucible, 
traversed by a stream of dry hydrochloric acid gas. Then 
after the crucible has been cooled and the hydrochloric acid 
replaced by dry air, the crucible is unsealed and the silicate 
quickly introduced. In this operation some moisture doubt- 
less gets in. Before heating again, dry indifferent gas is 
introduced. The crucible is heated for a number of days, the 
entering and exit tubes being guarded by driers. When the 
operation is completed, the excess of calcium chloride is 
removed by water. The product is usually in the form of 
transparent crystals of rhombic habit which sometimes attain 
to dimensions of several millimeters. The planes are often 
slightly coated with a thin film of what is probably calcium 
hydroxide, which doubtless comes from the decomposition of 
calcium chloride by the water vapor which could not be 
entirely excluded. It gives a strong alkaline reaction and may 
be readily removed with a little dilute HC1. Analysis shows 
that no lime is dissolved by the crystals, though they are not 
quite free from chlorine. 

* Comptes Eendus, lxvii, 43, 1868. fThis Journal, xxii, 389, 1906. 



14 Allen , etc. — Diopside and its delations 

Cal. for 

CaSi0 3 , 

Found MgSi0 3 

SiO„ _ 55-59 55*62 

CaO . 25-73 25-83 

MgO 18-61 18-55 

Fe 2 3 etc. -18 

CI trace 



100-11 100-00 

Several products made in this way were united and tested 
in a mixture of methylene iodide and benzene to see if all was 
of uniform specific gravity. A small fraction, containing some 
of the larger crystals, floated, while the rest sank. These crys- 
tals seemed to differ from the rest only in porosity. The 
remainder were then divided into two fractions of only slightly 
different density and the specific gravity of each was deter- 
mined by the pycnometer. 



■ ._. . , . Sp. gr. at 25° 

1. Lighter traction ' ° 

Jti \j at 25 


= 3-270 


2. Heavier " " 


3-275 


3. Natural diopside 




from Ham Island, 




Alaska, 


3-268 



An idea of the expansion of the diopside in the process of 
melting may be obtained by comparing this constant with the 
specific gravity of the glass of the same composition : 

2-830. 



HO at 25 c 



The description and angular measurements of these crystals 
are to be found in Part II of this paper. 

Melting Point Curve from J^6'3 — 100 per cent MgSiO s . — 
Referring again to fig. 1, it is seen that beyond 46*3 per 
cent xvIgSi0 3 , the addition of it lowers the melting point 
gradually but very slightly, until about 68 per cent has been 
reached. Near this point there is an abrupt rise, the curve 
finally terminating at 1524°, the melting point of magnesium 
silicate. On the first branch of the curve (46*3-68 per 
cent) mix crystals of magnesium silicate in diopside sepa- 
rate. This conclusion was tentatively reached in the prelimin- 
ary thermal work when it was found that the addition of a 
large percentage of magnesium silicate lowered the melting 
point of diopside very little, while the mixtures continued to 
show but one absorption of heat. The thermal evidence was 
subsequently confirmed by the specific volume curve, and by a 
very elaborate microscopic study. 



to Calcium and Magnesium Metasilicates. 15 

Solid solution determined by microscopic homogeneity ex- 
tends as far as 66'5 per cent MgSiO,: 33*5 per cent CaSi0 3 . 
The solid phases separating on the branch of the curve DE 
were found to be the mix-crystal just mentioned, and 
«-MgSi0 3 in all cases. All the mixtures from about 68-95 
per cent MgSi0 3 showed an absorption of heat at 1375°. This 
is evidently a eutectic line and there is no doubt that we have 
Jhere a case of Roozeboom's mix-crystal Type V,* where each 
of two mix-crystals lowers the melting point of the other to a 
eutectic. Points on this line to the left of the eutectic point 
could not be located since the liquidus curve lies only 2° to 3° 
higher up. The eutectic composition is about 68 per cent 
MgSi0 3 : 32 per cent CaSi0 3 and consists therefore of about 

32* 

_ — = 95*5 per cent of the diopside mix-crystal and about 4*5 

Oo'D . 

per cent of free a-MgSi0 3 ( in which about 2 per cent CaSi0 3 
or 3*7 per cent of diopside is dissolved ). 

Inversion of MgSiO y — About 1365° pure magnesium silicate 
undergoes a reversible change into an orthorhombic form. 
The heat of transformation is small, and apparently the change 
does not all happen at once, but extends over quite an inter- 
val of temperature. This is the only satisfactory way of 
accounting for the fact that while a small absorption of heat 
is observed in most of the mixtures of magnesium silicate from 
70 per cent on, in the pure substance and in the mixtures near 
100 per cent MgSi0 3 it was entirety overlooked. The exist- 
ence of a form stable above 1365° was proved by crystallizing 
a melt near the melting point, and then suddenly chilling the 
crystals in water. Thus the form was instantly brought into 
a temperature region where viscosity was great enough to 
prevent an inversion. The inversion temperature was deter- 
mined approximately as follows : One tenth of a gram of the 
substance was placed in a small platinum tray shaped from a 
piece of platinum foil. This was suspended by a fine plati- 
num wire inside the furnace. After the material was melted 
and crystallized near the melting point, the temperature was 
lowered to a given temperature and held there about a half 
hour. The furnace in which the heating was done was 
designed for calorimetric purposes, and had a removable 
bottom which was swung aside at the proper moment, 
when the suspension wire was instantly melted by an electric 
current, allowing the charge to drop into a basin of water. 
The results were as follows: 

*Zeitschr. f. Phys.-Chem., xxx, 403. 



16 Allen, etc. — Diopside and its Relations 

1 Chilled from an initial temperature of 1346°; all orthorhom- 

bic crystals (a-form). 

2 Chilled from an initial temperature of 1337°; all the crystals 

were orthorhombic. 

3 Chilled from an initial temperature of 1324°; all monoclinic 

crystals (/?-form). 

4 Chilled again from 1324° with the same result. 

After the existence of the a-form had been established, the* 
inversion point was repeatedly sought for by the Frankenheim 
method and a very small minimum between 1405° and 1435° 
was found in many cases bnt by no means in all, and the point 
was so small that one might well have hesitated to interpret it 
as an inversion, had other evidence been lacking. Taken in 
connection with the sharp heat change which occurs in the 
mixtures (see below) at about 1365°, it appears that the inver- 
sion point is overstepped about. 40° in both directions, which 
is in perfect accord with the sluggish behavior of solid silicates. 
The inversion- point in pure magnesium silicate is represented 
in the diagram (fig. 1) as falling somewhat as it dissolves cal- 
cium silicate, because this is the relation which commonly 
holds, but it is manifestly impossible to settle the question at 
present. 

a-Magnesium Silicate. — This form, which has already been 
described, crystallizes in the orthorhombic system in equant 
crystals on short doubly terminated prisms which resemble 
forsterite in habit, index of refraction and birefringence. 
Its specific gravity, judging from its index of refraction, varies 
little from that of /3-magnesium silicate (3*192). It was found, 
by applying the floating method to a single small crystal, to 
be about 3*16. Fortunately, in several instances a few well- 
formed, separately developed crystals were found on the sur- 
face of charges of magnesium silicate which had been melted 
and crystallized in the furnace, though usually all was com- 
pletely inverted into the magnesium pyroxene. The measure- 
ments of these crystals wdiich are recorded in Part II of this 
paper prove conclusively that this is a form entirely distinct 
from forsterite or enstatite. 

Inversion in the Mixtures. — The inversion line is traced in 
all the mixtures from 68 per cent to 98 per cent magnesium 
silicate and varies little from 1365.° Some typical curves 
showing both the inversion and eutectic points are plotted in 
figs. 4 and 5, curves II, V, VI, VII, and VIII. The signifi- 
cant feature of these curves is the gradual diminution of the 
heat absorption at the latter (eutectic) temperature and the 
gradual increase of that at the former (inversion), as we pass 
toward 100 per cent magnesium silicate. Thus in the 70 per 
cent mixture the eutectic melting is large, while in the 95 per 



to Calcium and Magnesium Metasilicates. 
Fig. 4. 



17 



























































/ 


































































/ 
































































/ 












i 






















































/ 












1390° < 




















































/ 


































































/ 


























































■s 
























































































1370° 














































-*■ 






















































































1360° 














\ 














































































* -i 


►■ 














































y 


<r 


























































• 


s 


r"' 




























































< 


f 






, 





























































































































Temperature-time curves showing thermal behavior of mixtures rich in MgSi0 3 . 
Curve I-I, 60 % MgSi0 8 . II-II, ?0 % MgSi0 3 . 



Fig. 5. 



1410 c 



1400° 



1390° 



1380° 



1370° 



1350° 



1340° 




Temperature-time curves showing thermal behavior of mixture rich in MgSi0 3 . 

Curve III. 65 % MgSi0 3 . 

IV. 75 % MgSi0 3 crystallized below 1300° by heating the glass. 

V. To % " " not far below 1365°. 

VI. 80 £ " " " " ." " 

VII. 90 £ " " l < " " 

VIII. 95 £ " " " " " 

Am. Jour. Sci. — Fourth Series, Vol. XXVII, No. 157.— January, 1909. 
2 



18 Allen, etc. — Diopside and its Relations 

cent it is minute and in the 98 per cent can be no longer 
detected ; while the inversion, small in the 70 per cent mixture, 
is much greater in the 90 per cent and 95 per cent mixtures. 
The 98 per cent mixture, where solid solution of calcium sili- 
cate begins, shows the inversion plainly but not so markedly as 
the 95 per cent mixture. It evidently partakes of the char- 
acter of the pure silicate, which is more resistant to rapid 
change. Dissolved calcium silicate therefore facilitates the 
inversion of magnesium silicate, while, as we shall presently 
see (page 20), dissolved magnesium silicate hinders the inver- 
sion of calcium silicate. It is not strange that the properties 
of a substance should be modified by solid solution, though it is 
evident that the direction of such a change cannot yet be pre- 
dicted, but why an excess of foreign solid should affect the 
inversion point of a substance is not clear. The influence of 
solid diopside in concentrating or sharpening the inversion of 
magnesium silicate, however, seems to be established. 

Curves I and III show the behavior of materials which con- 
tain no free magnesium silicate and therefore show no inversion. 
Curve IV shows what happens when mixtures are chilled to 
glass and then heated. Crystallization then leads to enstatite, 
which evolves heat slowly over a long range of temperature. 
Since this evolution covers the region above 1300°, the small 
absorption at 1365° is effectually hidden, where, as explained, 
the inversion seems to be lengthened out. 

Solid Solutions. — The system CaSi0 3 -MgSi0 3 contains six 
different series of solid solutions, in only two of which is there 
more than a few per cent of the smaller constituent. The 
limit of solubility was determined by the thermoelement, by 
the microscope or by both when possible. 

So far as thermal tests are concerned, the results refer of 
course to the presence or absence of a eutectic melting, i. e., 
the solubility at the eutectic temperature is approximately 
determined. The microscope, on the other hand, makes its 
determinations at ordinary temperatures, where presumably the 
solubility is generally greater. In almost all cases the solu- 
bility as determined by the microscope is a little higher, but 
how far this is a real difference and how far it is due to a differ- 
ence in the delicacy of the methods, it is unsafe to say on 
account of the uncertainty in establishing equilibrium in these 
solid silicate solutions ; in other words, it is impossible to say 
whether a given solution is saturated or not. In the diagram, 
fig. 1, we have drawn the lines straight with a single exception, 
where we have more evidence that solubility increases with 
falling temperature. 

1. Diopside in a-MgSi0 3 . On account of the difficulties 
involved, this series was not investigated. 



to Calcium and Magnesium Metasilicates. 19 

2. Diopside in /3-MgSi0 3 . The microscope detected inho- 
mogeneity in the 3 per cent CaSi0 3 mixture, but none in the 2 
per cent. No eutectic was observed in the latter. We may 
therefore put the solubility as 2 per cent CaSi0 3 , or 3*7 per 
cent diopside. 

3. MgSi0 3 in diopside. The microscope places the limit of 
solubility at about 66*5 per cent MgSi0 3 : 33*5 per cent CaSi0 3 , 
i. e., the quantity of diopside in the saturated mix-crystal 
would be 33-5/53*7 = 63*4 per cent. In other words, diopside 
is capable of dissolving 3 T 6/62 "4 = more than 60 per cent of its 
own weight of magnesium silicate. This remarkable series of 
mix-crystals strongly resembles diopside. The optical work 
described in detail in Part II shows that the extinction angle 
and the optic axial angle both fall about \° for each 
additional per cent of magnesium silicate. The specific 
volume curve (fig. 9) shows that the solution is attended by 
expansion. 

4. Calcium silicate in diopside. The microscope detected 
inhomogeneity at 44*5 per cent MgSi0 3 while the thermoelement 
detected a plain eutectic in the 45 per cent MgSi0 3 . The 
limit of solubility is doubtless small, certainly less than 100 — 
45/46-3 = 3 per cent CaSiO,. 

5. Diopside in a-CaSi0 3 (pseudo-wollastonite). Mixtures 
containing as much as 3 per cent MgSi0 3 plainly showed 
inhomogeneity when examined by the microscope. The 
inhomogeneity took the form of irregular bands, irregularly 
distributed, which showed a distinctly lower birefringence 
than the rest. The 2 per cent MgSi0 3 solution showed traces 
of the above structure and gave a plain eutectic when examined 
thermally. The 1 per cent solution is microscopically homo- 
geneous and gives no more than a suspicion of a eutectic. The 
error will be slight if we put the limit of solubility at about 
1-2 per cent MgSi0 3 , or in round numbers, 3-4 per cent of 
diopside. 

6. Diopside in /3-CaSi0 3 (wollastonite). In this case a 
thermal test is obviously useless. The microscopic analysis 
showed that solution ceased at about 17 per cent diopside (8 per 
cent MgSi0 3 ). Crystallization took place at about 1058°. The 
crystals of diopside and wollastonite are both monoclinic and 
the latter resembles diopside more closely than pseudo-wollas- 
tonite, so that it is not surprising that wollastonite should dis- 
solve diopside in large quantity. 

These solutions showed a very interesting behavior when 
they were heated. To understand this clearly, it should be 
remembered that the eutectic between pseudo-wollastonite and 
diopside melts at about 1348°, while wollastonite has an inver- 
sion point at about 1190°. Upon heating crystals of wollaston- 



20 Allen , etc. — Diopside and its Relations 

ite carrying 1 per cent MgSi0 3 (2*1 per cent diopside) for one 
hour at 1221°, no change was apparent. Heated again for two 
hours at 1245°, the inversion was slight. The inversion point 
of wollastonite appears, therefore, to be raised 40°-50° by the 
solution of only 2 per cent diopside. (See p. 18.) 

Since time is an important factor in sluggish changes, a 
direct comparison was made between pure wollastonite and two 
mix-crystal preparations of this series by heating all three in 
the same furnace for the same length of time, viz., 1 hour. 
The temperature ranged from 1257 6 ~to 1263°, i. e., about 65° 
above the inversion point of wollastonite. The wollastonite 
was completely inverted, the 2*1 per cent solution slightly 
inverted, while in the 4\3 per cent solution (2 per cent MgSi0 3 ) 
it was doubtful if any change at all had taken place. The two 
solid solutions were now returned to the furnace and held an 
hour longer between 1273° and 1300°, about a hundred degrees 
above the inversion point of pure calcium silicate. The weaker 
solution was now found to be much changed, the stronger one 
less so. The 8 per cent solution of MgSi0 3 , containing 17'3 
per cent diopside, was heated for an hour at 1278° to 1280°, 
90° above the inversion point of calcium silicate. A careful 
microscopic examination of the crystals then showed a con- 
siderable change in their appearance ; a new product had sepa- 
rated but it did not show the characteristics of pseudo-wollas- 
tonite. Such optical properties as could be measured in fine- 
grained material (index of refraction, birefringence) agreed 
with diopside. This indicates that the solubility of diopside 
in wollastonite is greater at lower temperatures where the 
crystallization occurred, or perhaps that the solid solutions, 
being formed by rapid crystallization, were supersaturated. In 
either case the excess separates when the solution is heated to 
higher temperature. 

Heated an hour longer at a temperature of 1298°-1303°, 
more diopside separated, but the signs of inversion were still 
doubtful. Again, the crystals were returned to the furnace 
and the heating continued another hour at 1327°-1343°. 
This time inversion was evident. These experiments show 
either a marked rise in the inversion temperature of wollas- 
tonite or else a great increase in molecular sluggishness caused 
by the dissolved diopside. 

A brief discussion will make it clear under what conditions 
an inversion point may be raised and thus help to decide 
whether we have a real rise in the inversion temperature or 
not. By a thermodynamic method Beckman^ has shown that 
the freezing or inversion point of a substance is changed by 

* Ostwald's Lehrbuch der Cliemie, vol. 2, pt. 2, pp. 38, 68. 



to Calcium and 



ium Metasilicates. 



21 



Table III. 

Inversion Temperature of Wollastonite- D topside Mix- Crystals. 

(Inversion point of wollastonite, 1190°.) 



No. 

of 
Exper. 



Composition 



T ^ e Temper- 
Heating 



ature 



1 \% MgSiOs (2-l£ diopside) 

2 1% MgSiOs (2"1£ diopside) 

Continuation of Exper. 1 
0£ MsrSiOa 



1 hr. 
2hr. 



lhr. 



3 1% MgSi0 3 {%'!% diopside) 
2 r c MgSi0 3 (4-3^ diopside) 

. \ \% MgSiOs (Continuation ) 1 , 

4 \ 2% MgSi0 3 of Exper. 3) f i nr> 

5 8% MgSiOs (17 '3^ diopside) 1 hr. 



1221 c 
1245 c 



1257-1263° 

1273-1300° 

1278-1280* 



6 8% MgSi0 3 (17 -3£ diopside) 

(Continuation of Exper. 5) 1 hr. 1298-1303 c 

7 Continuation of Exper. (5 1 hr. 1327-1343 c 

Fig. 6. 



Eesults. 

No change. 
Inversion slight. 

Completely 

inverted. 
Slightly " 
Doubtful 
Largely inverted 
Much less ' ' 
Separation of 
diopside. No 
inversion. 

Inversion 

doubtful. 
" evident. 




Fig. 6 shows under what conditions an inversion point remains unaltered 
after the formation of a solid solution. 

the solution of another substance according to the equation 

02 T 2 

A =— — (C. — C„) where A = the depression in the tem- 
e 

perature, T = the absolute temperature of the inversion point in 
the pure substance, I — its latent heat of fusion, C, — the con- 
centration of the solution above the inversion point and C 2 = 
the concentration of the solution below the point. 



22 Allen, etc. — Diopside and its Relations 

This formula holds approximately for concentrated solutions. 
When Cj = C 2 , i. e., when the concentration of the solid solu- 
tion is not changed by the process of inversion, there will 
evidently be no change in the inversion point; when C X >C W 
the temperature will fall, but if C 2 , the concentration below 
the freezing point, is greater, then A will have the opposite 
sign, and the freezing point will be higher than that of the 
pure substance. 

The same conclusion is reached by a graphic method.* In 
fig. 6 let AB represent the vapor pressure curve of the pure 
solid below the inversion point, BC the vapor pressure above 
it, and T the inversion temperature. Now if this solid forms 
a solid solution, the vapor pressure of the former will be 
lowered according to the concentration of the solution. Sup- 
pose that this is the same above and below the inversion point, 
and that the vapor pressures are lowered to the same degree in 
both. It is evident that the new curves DE and EF will inter- 
sect at the same temperature and the inversion temperature is 
therefore unchanged. 

In fig. 7 let us suppose that the concentrations of the two 
solid solutions are unequal, the one below the inversion point 
being the more dilute. The curve AB will be lowered to DE 
and BC will fall to EF by reason of the greater concentration 
of the second solution. DE and EF now intersect at E, at a 
temperature lower than T . By similar reasoning we conclude 
that when the solid solution below the inversion point is the 
more concentrated the inversion point will be raised. (See 
fig 8.) Apparently the lower concentrations of diopside in 
wollastonite remain the same when the mix-crystals invert, 
for about 1-2 per cent MgSi0 3 is dissolved by the pseudo- 
wollastonite. In these cases, "therefore, the inversion point 
theoretically should not change. These solutions, however, are 
unquestionably less changed at the same temperature than 
wollastonite is. We therefore conclude that the solution of 
diopside has increased the intermolecular friction of the crys- 
tals. It is also possible that these solutions of wollastonite- 
diopside which are more concentrated than 1-2 per cent 
MgSiOg really have a higher inversion point than 1190°. 
If the crystals saturated at this temperature remain more con- 
centrated in diopside than the pseudo-wollastonite is, this must 
be true. A decision cannot be reached until we have some 
sure method of establishing equilibrium. Day and Shepherd 
found that solid solutions of lime or silica in calcium metasili- 
cate inverted to wollastonite on cooling. Since the pure rneta- 
silicate does not behave so, we naturally conclude that the 
internal friction was lessened by the lime or silica. Magnesium 

* Bodlander, Neues Jahrb. Min., Beilage Bd. xii, p. 52. 



to Calcium and Magnesium Metasilicates. 



23 



metasilicate has no such influence. The 4*3 per cent of diop- 
side in pseudo-wollastonite was cooled from 1209° to 1050° 
during a period of 1J hours without producing any change 
in the crystal form. In the diagram, fig. 1, we have drawn a 



Fig. 7. 




Ti T 



Fig. 7 shows under what conditions the inversion point of a solid is 
depressed after the formation of a solid solution. 

dotted line through the points where this series of solutions 
was actually observed to invert, but it must not be accepted as 
a true inversion line. The inertia of these mix-crystals of wol- 
lastonite causes them to exhibit a curious and variable behavior 



Fig. 8. 




1 T i 



Fig. 8 shows under what conditions the inversion point is raised after 
the formation of a solid solution. 



24 Allen, etc. — Diopside and its delations 

when heated, which in the beginning of the work was quite 
confusing. The normal behavior of a solution of wollaston- 
ite containing more than 2-4 per cent of diopside, (the 
limit of solubility of the latter in pseudo-wollastonite) would 
be as follows : First, an inversion should occur somewhat above 
1190°, giving crystals of pseudo-wollastonite saturated with 
diopside, and an excess of free diopside. At 1348° some 
eutectic melting would be noted, and finally, when the melting 
point curve is reached, another thermal point would be found. 
Under ordinary experimental conditions, however, where the 
rate of heating is about 3° per minute, solutions containing 5 
per cent MgSi0 3 (10'8 per cent diopside) showed no melting at 
the eutectic temperature. Evidently no inversion had taken 
place. When the rate of heating was considerably slower a 
slight eutectic melting was noted, while if the crystals were 
previously held for some time at 1360°, there was a strong 
absorption of heat at the eutectic temperature. In the 8 per 
cent MgSi0 3 solution (1T'3 per cent diopside) the eutectic failed 
in one instance, i. e., there was no melting at the eutectic 
temperature. Mixtures of saturated mix-crystals of wollaston- 
ite-diopside give, when the heating is not too slow, three points, 
the eutectic at about 1348°, a further melting accompanying 
the inversion at a higher temperature,* and finally, the point 
on the melting point curve where all becomes liquid. Thus 
the 10 per cent mixture (containing about 4 per cent of free 
diopside) showed heat absorptions at 1340°, 1377° and 1451°. 
The 8 per cent solution showed a similar behavior, indicating 
that some inversion may have occurred below the eutectic points. 
Another explanation is perhaps more probable, viz. : that the 
lowest point may be due to an unstable eutectic between wollas- 
tonite and diopside, since the latter was found to separate from 
the more concentrated solutions at the higher temperatures. 
If so the point lies very near the pseudo-wollastonite diopside 
eutectic. A fact that seems to favor the explanation is that 
the 10 per cent solution always gives this lowest point although 
it has been proved that the more concentrated solutions invert 
with greater difficulty. Moreover, when the 28 per cent solu- 
tion was crystallized below 1190° and therefore contained no 
pseudo-wollastonite (a conclusion also verified by the micro- 
scope), all melted at 1358° as usual. 

Specific-volume curve. — In 1890 Retgersf stated clearly two 
arguments to prove that diopside was a chemical compound in 
distinction from a mix-crystal. The first was that the minerals 
in nature which contain the metasilicates of calcium and 

* Since the solution of the diopside in the wollastonite crystals prevents 
partly or wholly the eutectic melting at the proper temperature, this melting 
will at once occur when the crystals are inverted. 

f Ann. Ecole Polytech. de Delft, iv, p. 186, 1890. 



to Calcium and Magnesium Metasilicates. 25 

magnesium vary comparatively little from the compositions 
CaSi0 3 , MgSi0 3 and CaMgSi 2 6 ; and the second, that the 
specific volume of diopside could not be calculated additively 
from the volumes of the constituents. Retgers's method of 
solving the question of isomorphism or isodimorphism between 
two substances is well known. It consists in the preparation 
of a suitable series of mix-crystals of the two substances and a 
study of the relation which their specific volumes bear to one 
another. He proved by many examples that the specific vol- 
umes of isomorphous mixtures (as he defined them), when 
plotted as a function of the composition, form a straight line. 
In the paper quoted above, Retgers said that this would be 
the best way to prove whether calcium and magnesium sili- 
cates form a double salt or are isodimorphous, if their mixtures 
could only be crystallized in sufficiently large individuals for 
specific gravity determinations. He used the floating method, 
which is not adapted for very small particles, and he empha- 
sized the importance of making sure that the material is both 
physically and chemically homogeneous. He therefore used 
only transparent individuals for fear that aggregates might 
contain some foreign material which would escape optical 
detection. It has been shown in this laboratory that the spe- 
cific gravities of mineral powders, if not too fine (100-120 mesh), 
can be determined with a degree of accuracy (±001 for 
substances of the gravity of 3) very nearly as great as those 
obtained by Retgers's method. Of course, the particles should 
be free from air bubbles or vacua, and it must be admitted 
that powders require a very careful microscopic investigation 
to decide this point. Mixtures of calcium and magnesium 
silicate generally show a certain amount of " dustiness" due 
to very minute inclusions, or more probably to vacua. These 
are more numerous in the mixtures which are rich in mag- 
nesia (70-97 per cent), but not in the pure magnesium 
silicate itself. When large masses of material (100 grams) are 
crystallized slowly, the density is greater and the microscope 
shows that the vacua are fewer and smaller. Although the 
specific gravities of the mixtures crystallized in this way are 
still too low, we judged that they would probably be approx- 
imately comparable among themselves, and this conviction 
has been justified by experiment. 

The specific- volume curve (fig. 9) plainly consists of three 
branches. AJB is the locus of the volumes of mechanical mix- 
tures of the pseudo-wollastonite and diopside (leaving out of 
the question the small mutual solubility). Independently of 
microscopic or thermal evidence, it would, of course, be 
impossible to say whether this line indicated a series of 
mechanical mixtures or a series of mix-crystals between two 



26 



Allen, etc.- 



■Diopside and- its Relations 



isomorphous substances. In either case the volumes would 
form a straight line between the two constituents. DC is 
made up of diopside mix-crystals with magnesium silicate. 
The limit of the solubility of the latter in diopside is at about 
70 to 72 per cent MgSi0 3 as determined by the specific volume 



Table IV. 

Specific Gravities and Specific Volumes of Mixtures of 
CaSi0 3 and MgSi0 3 . 



Ofo J 


tfftS 


8 


u 


10 


a 


20 


it 


30 


a 


40 


a 


45 


it 


46-3 


tt 


47*5 


a 


50 


a 


55 


a 


60 


a 


65 


it 


70 


a 


72 


a 


75 


a 


80 


a 


85 


a 


90 


a 


95 


a 


100 


a 



3p. gr. 


Sp. v. 


2-912 


•3434 


2-965 


•3373 


2-947 


•3362 


3-046 


•3283 


3-111 


•3215 


3 201 


•3125 


3-229 


3096 


3-237 


•3089 


3-236 


•3090 


3-241 


•3<»86 


3-246 


•3081 


3-245 


•3<!82 


3-255 


•3()72 


3-241 


•3086 


3-241 




3-229 


•3096 


3-211 


•3115 


3-212 




3-205 


•3120 


3-198 


•3127 


3-196 


•3129 


3-196 


•3127 


3-194 


•3130 


3*198 


•3127 


3-192 


•3132 


3-188 


•3137 


3-181 


•3140 


3-193 


•3132 



Where two different numbers are given, they belong to different prepara- 
tions. 

curve. The microscope sets the limit at about 66'5 per cent. 
The discrepancy is probably due to experimental error in the 
specific gravity determinations caused by the presence of bub- 
bles in the grains. An inspection of the curve shows that the 
volumes of these solid solutions all lie above a line joining B 
and D. They are, therefore, greater than the volumes calcu- 



to Calcium and Magnesium Metasilicates. 



27 



lated on the assumption of a purely additive relation. A sim- 
ilar expansion is known in other cases.* CD contains the 
volumes of mixtures of magnesium silicate and the diopsicle 
mix-crystals. The minimum B, indicating a compound, falls at 
about 49 per cent MgSi0 3 instead of 46*3 per cent, which is 
demanded by the formula CaSi0 3 .MgSiG 3 . This is because a 



Fig. 





A 
























\ 
























\ 
























\ 


• 
\ 
























\ 
























\ 
























\ 


\ 
























\ 
























\ 
























\ 


\ 
























\ 
















•3150 








\ 

\ 
























\ 


\ 






pr*- 1 




> 


D 


■3100 










\ 




/ 


>C 


















\ 


/ 












•3050 










} 


/ 













10 20 
CaSi0 3 



40 50 60 70 
Weight per cent. 



110 
MgSiO a 



melt of the latter composition forms a crystalline mass which 
always appears to contain more bubbles than the other compo- 
sitions in its immediate neighborhood, so that the density of 
the crystals is not only absolutely but also relatively too low. 
It will be remembered in this connection that the diopside 
which was crystallized from calcium chloride had a specific 
gravity of 3'275, while that which was crystallized from a melt 
of the composition CaMgSi 2 G 6 had a density of only 3*24. 

*E. S. Shepherd, Journ. Phys. Chem., viii, 245, 1904. 



28 Allen, etc. — Diopside and its Relations 

We made a number of experiments witli the intention of 
finding whether the mix-crystals of diopside and magnesium 
silicate could be crystallized again from calcium chloride, or if 
not, what change in composition they would show with varying 
quantities of the chloride, but the products obtained were not 
only not homogeneous, but the crystals were too small to sepa- 
rate from one another and the microscope was unable to 
identify them. One difficulty in these experiments was the 
impossibility of entirely excluding water from the apparatus. 
Its reaction with the calcium chloride, of course, formed some 
free lime, which was dissolved by the silicate. Pure diopside, 
as we have seen, does not dissolve lime, but magnesium silicate 
does. Three grams of the latter, containing only 0*12 per cent 
of lime, was crystallized* from calcium chloride and analyzed 
after the excess of the reagent was removed. It now contained 
3*04 per cent of lime, an increase of 2*92 per cent. 

Part II. Optical Study, by Feed. Eugene Weight and 
Espee S. Laesen. 

In the foregoing pages the theoretical aspects of the Ca-Mg- 
metasilicate problem have been treated at length, evidence 
from all sources, chemical, physical, optical and crystallo- 
graphic, having been brought to bear on its solution. In this 
general presentation of the problem, however, only the more 
important and decisive optical and crystallographic data have 
been made use of, their detailed tabulation having been 
reserved for a separate section. In the following paragraphs 
those details which are still lacking are listed, and in order to 
avoid repetition, general theoretical considerations have been 
avoided so far as possible. In the attack on the present prob- 
lem the effort has been made from the very first to combine 
the evidence from all viewpoints and to test each conclusion 
by such evidence. It has been found that by this method the 
constant interchange of ideas and the discussion of the details 
of the problem have tended greatly to improve and to 
strengthen the final result. 

For the sake of convenience the optical and crystallographic 
features of the three compounds of this series will be considered 
first, after which will follow the particular features of the 
intermediate preparations. 

Calcium Metasilicate. — The two enantiotropic forms of 
this compound, wollastonite and pseudo-wollastonite, have 
already been described in detail in this Journalf and the 
evidence need not be repeated at this point. Since the publi- 

* The crystals were small but well-developed. 
f This Journal, xxi, 103-108, 1906. 



to Calcium and Magnesium Metasilieates. 29 

cation of the above paper, however, better facilities for refrac- 
tive index determinations have been acquired and the refractive 
indices of the two compounds have been redetermined. The 
measurements were made on polished plates of the crystalline 
aggregate, experience having taught that even under such 
conditions it is possible to determine 7 and a in sodium light 
with the reducing attachment of the Abbe-Pulfrich total 
refractometer, while ft can also be ascertained if the individual 
grains are sufficiently large. For wollastonite the new values 
are, y^ = l'632±-002 ; /3 Na = l--628=b"003 ; a Na = 1-616 + 002 ; 
ry — a— 016, 7 — j3 = -004:, j3— a — '012. The refractive indices 
of pseudo-wollastonite are : 7 Na =l"650=b'002 ; a Na =l*609±003 ; 
birefringence 7 — a = *041. The birefringence of both wollas- 
tonite and pseudo-wollastonite was furthermore checked by 
direct measurement in the thin section ; 7— a for wollastonite 
being *014 and for pseudo-wollastonite *043.* 

Magnesium Metasilicate. — Like the calcium metasilicate, 
this compound has also been described in a special paper in 
which the optical characteristics are considered together with 
the other properties. Four different forms or phases were 
there mentioned bearing mono tropic relations to each other, 
the monoclinic Mg-pyroxene being the one stable form. In 
the course of the past winter, however, still another phase has 
been discovered, orthorhombic in symmetry and in general 
aspect and development not unlike that of olivine crystals. 











Table V. 








No. 


Letter 


Symbol 


Miller 


4 


1 


P 


1 


c 


(?) 





001 


.. 




2° 00' 


2 


a 




00 


100 


90° 


00 


90 00 


3 


b 




0oo 


010 





00 


a 


4 


m 




00 


110 


44 


04 


a 


5 


n 




00 2 


120 


23 


07 


a 


6 


I 


(?) 


»t 


250 


17 


56 


a 


7 


k 


(?) 


3oo 


310 


70 


44 


a 


8 


r 


(?) 


2oo 


210 


66 


54 


a 


9 


s 




11 


111 


43 


20 


38 58 


10 


e 




12 


121 


24 


58 


52 11 


11 


P 




10 


101 


90 


00 


29 09 


12 







1 


111 


46 


48 


40 21 


13 


i 




12 


121 


28 


30 


53 10 


14 




(?) 


i<> 


103 


90 


0*0 


11 55 


15 




(?) 


*o 


103 


90 


00 


12 55 



* Opportunity may liere be improved to correct several of the optical data 
in the paper on the Lime-Silica Series of Minerals (this Jour. , xxii, 293-302, 
1906). At top line of p. 297 read : a = 1-609 ± '003, 7 = 1 -620 ± '002, instead 
of the values given ; on p. 298, 11 lines from the top, e = 1'554± '002 and 
w = l-544±-002; on p. 299, 8 lines from bottom, 1-585 and 1*621; and 2 
lines from bottom, a = 1-590 ; on last line, p. 299, read "019 instead of "025. 



30 Allen, etc. — Diopside and its delations 

The optical data of the monoclinic Mg-pyroxene form 
given in the above article can now be supplemented by more 
accurate figures in several instances ; and at the same time 
several errors in the crystallographic data can be rectified. 
The /3-MgSi0 3 is monoclinic and the observed form and 
measured angles are listed in Table V above.* 

The reflexion signals from many of the faces were multiple 
and wide variations in the angles occur. The best average 
ratios from these values are about 

p = -58 q = -60 e = -046 fx = 87°26 ; or 
a : b : c= 1*03 : 1 : 0-60 ; /3 = 87°26' ; 

values which are closely similar to those for enstatite but less 
so for diopside. The variations are not such, however, as to 
preclude isomorphic relations between the two. The plane of 
the optic axes lies normal to the plane of symmetry and not in 
the plane of symmetry as in most pyroxenesf ; the bisectrix c 
is inclined to the vertical axis c, 21*8°. The optic axial angle 
was measured on the universal stage and by means of the two 
screw micrometer ocular. Care was taken to select favorable 
sections and the probable error in each case was not large. 
The average of all good determinations by both methods is 
2V = 53'5°±1°, or 2E = 96°. The overlapping of the twin- 
ning lamellae often causes abnormal variations in this optic 
axial angle and it proved a difficult matter to find suitable 
sections. Etch figures on the cleavage face were also obtained 
and will be considered later, together with the etch figures of 
the other members of this series. 

The a-MgSiOg, to which reference has already been made, 
crystallizes readily and is obtained by quenching the crystal- 
lized melt from temperatures above 1365° to prevent its 
inversion to the /3-form. Once obtained it can be held for 
apparently an indefinite period at ordinary temperatures with- 
out inversion to the more stable /3-form. The crystals are 
orthorhombic in symmetry and in certain positions bear strong 
resemblance to the characteristic crystal habit of olivine. On 
the surface of a melt held at 1510° and then chilled rapidly, 
several crystals, water-clear and sharply bound crystallographi- 
cally, occurred, and three of them (1 X '5 X '2 mm ) were measured 

* Compare with Table II on page 393, vol. xxii, 1906, this Journal, in which 
several letters were unfortunately transposed, although the relations are 
correctly represented in the gnomonic projection plat on the same page. 

f The relations of the different pyroxenes, particularly of the magnesium 
iron group, have been recently discussed in an interesting paper by W. Wahl 
(Die Enstatit-augite, Tschermak's Miner. Petrogr. Mittheil., xxvi, 1-131, 
1907), who proposes the name clino-enstatite for the /?-MgSi0 3 or magnesium 
pyroxene. The suggestion is a good one and may well be adopted, the two 
latter terms being long and cumbersome. 



to Calcium and Magnesium Metasilicates. 31 

on the goniometer. The reflexion signals were not of the best 
and the angles of the table can be considered only approxi- 
mately correct, an error of ±15' being easily possible. 









Table VI* 






No. 


Letter 


Miller 


Symbol 





p 


1 


I 


010 


Ooo 


0°03 / 


90°00' 


2 


m 


110 


00 


40 08 


90 00 


3 


k 


Oil 


01 


01 


25 09 



From the angles the crystallographic constants can be cal- 
culated : j? = 0*40 q = 0*47 
or a : h : c = 1-19 : 1 : 0'47. 

The crystals are often tabular and prismatic in shape 
after 010 as indicated in fig. 10. In other cases the prism 
zone is less prominent and the crystals are of 
eqnant development. The forms b, m, and Jc Fl &- 10. 
were observed on all three crystals. Cleavage after 
100, good. 

The plane of the optic axes is the cleavage plane 
100 and the acute bisectrix is c. The optical orien- 
tation is, therefore, a = ft ; b = c ; and optical char- 
acter + . The refractive indices were determined 
by the immersion method in refractive liquids ; a 
= 1-641 ± -003 ; £ = 1-648 ±'003 ; 7 = l-663db'003 ; 
ry_ a =-022; 7— £ = '015; 0-a = -OO7. The birefringence 
7— /3 = "016 was furthermore measured directly on a tabular 
crystal 0*182 mm thick by use of the Babinet compensator. 
The optic axial angle is large and was measured with the 
two-screw micrometer ocular on two sections showing an 
optic axis in the field of vision. The average of the two 
values thus obtained (2Y = 59*5° and 61*0°) is about 2Y 
= 60-3° and 2E = 111°. The axial dispersion is fairly 
strong, 2 Vp > 2Yv. 

* The angles in the table are the averages of the different values obtained. 
These measurements were made on January 17, 1906, and since that time no 
suitable crystals for goniometric measurement have again been observed. 
Although the original notes stated definitely that these "crystal angles do 
not coincide with those of enstatite " while "the optical relations do not 
correspond with those of olivine," and this same form was observed at that 
time in at least seven different preparations, its importance was not realized 
and its presence was ascribed to impurity. In later experiments practically 
no chilling was done and not until the thermal data indicated to Dr. A. L. 
Day the presence of a high-temperature phase, enantiotropic to the first, 
were quenching experiments again resumed and with them the true signi- 
ficance of the a-MgSi0 3 became apparent. 



32 Allen, etc. — Diopside and its Relations 

In several of the larger crystals of the <x-MgSi0 3 a character- 
istic arrangement of inclusions and lines of growth was observed 
and strongly resembled the hourglass structure of certain 
pyroxenes, the hourglass portions of each crystal showing 
abnormal interference phenomena, due either to incipient 
changes into the/3-form or to peculiar intergrowths or possibly, 
but not probably, to strain phenomena. 

The a-form is readily distinguished from the /3-form by its 
lack of poly synthetic twinning, parallel extinction and stronger 
birefringence ; from olivine and enstatite by its cleavage and 
the position of the optic axial plane relative to the cleavage. 

It is of interest to note that the <x-MgSi0 3 , which is unstable 
below 1365° and cannot be obtained except under very special 
conditions, has not been observed in nature, thus establishing, 
as in the case of pseudo-wollastonite, a high temperature limit 
for the formation of certain minerals. 

Diopside. — This third compound of the series is an excellent 
crystallizer and can be formed in a number of different ways 
and at different temperatures. The best crystals were obtained 
by heating glass of the composition CaMgSi 2 6 in a flux of 
CaCl 2 in an atmosphere of dry HC1 at 1000° for one week. The 
crystals varied in size up to 2 mm in diameter, were water-clear 
and of simple crystallographic habit. Three crystals were 
measured on the goniometer with reducing attachment. The 
reflexion signals obtained were not of the best and the values 
of Table VII are the averages of the observed angles. A 
number of other crystals were selected and mounted on the 
goniometer, but the reflexion signals from their faces were 
often multiple and unsatisfactory, and not suited to improve 
the results already obtained from the three measured crystals. 











Table VII 


[. 














Artificial 


Diopside 


Natural 


Diopside 


No. 


Letter 


Miller 


Symbol 


<?> 


P 





P 


1 


b 


010 




00 


90 01 


00 


90 00 


2 


m 


110 




43 30 


90 "0 


43 33 


90 00 


3 


s 


11 1 


-1 


24 51 


33 + 


25 07 


33 04 


4 


X 


221 


-2 


35 12 


55 2<> 


35 22 


55 19 


5 


X(?) 


331 


-3 


39 17 


6tf 17 


38 19 


66 04 



For the sake of comparison, the angular values <j> and p for 
the same forms on natural diopside (Goldschmidt, Winkelta- 
bellen, p. 283) are included in this table. From these angles 
the average value of ft = "539; q = "568; e — *276, and \x == 
73° 59'; or a : b : e = 1-096 : 1 : '591. For natural diopside, 
j> o = 0-5390 ; q = 0-5670 ; e -- 0-2731 ; fi =-- 74° -09 ; ; or 




to Calcium and Magnesium Metasilicates. 33 

a\h:G~ 1*0931: : 1 : 0*5891. Both the angular values and the 
calculated erystallographic constants prove the close resem- 
blance of the artificial diopside crystals to the natural mineral, 
the differences observed being within the limits of error possible 
with the quality of reflection signals _, 11 

obtained from the artificial crystals. 
Except for the form X, which was observed 
only once, the crystal habit is simple 
and requires no comment. Twinning after 
100 is common and one of the three 
crystals measured was thus twinned, the 
twinning plane dividing the crystals into 
two nearly equal halves. Postsynthetic 
twinning after 100, however, occurred only 
rarely and is not characteristic. 

Prismatic cleavage after 110 is good ; in diopside crystallites 
from the melt indications of a parting after a dome or basal 
pinacoid face at an angle of about 66° with the prismatic 
cleavage cracks were also recorded. 

The refractive indices were measured on polished plates of 
the crystalline material on the total refractometer in sodium 
light : 

y Na = 1-694 ± -002 ; /? Xa = 1-671 ± '002 ; a Na = 1*664 ± -002 
y — a — -030 ; y — /3 = "023 ; f3-a = *007 

Direct determinations of the birefringence were also made in 
the thin sections with the result : y — a — "030 ± "002 (average 
of three measurements on good sections). The optic axial 
angle was measured on a number of different sections both 
with the universal stage and with the two-screw micrometer 
ocular. The average of four good determinations is : 
2Y = 59°*3 =t 1° ; 2E = 114°. The optical axial dispersion 
is weak, 2V/o> 2Y V . The plane of the optic axes is the plane 
of symmetry (010). 

Extinction angles were measured both on the clinopinacoid 
and the prism face 110. On 010 c : c = - 38°'5 =b 1°. On 110 
c : c = — 32°'9 ± 1°. The position of total extinction was 
ascertained in each case by use of the new bi-quartz wedge 
plate* and the values should be correct within 1°. Extinction 
angles were also measured for different faces in the prism 
zone. Suitable crystals, held in a specially constructed device, f 
were immersed in a liquid of the refractive index /3 and the 
extinction angle read for different angles of revolution of the 
crystal from its position of zero extinction when the ortho- 
pinacoid is normal to the axis of the microscope., 

* This Journal, xxvi, 377-379, 1908. f This Journal, xxvi, 388, 1908. 
Am. Jour. Sci.— Fourth Series, Vol. XXVII, No. 157.— January, 1909. 



34 



Allen, etc. — 1 


liopside and its 


Relation 


4>* 


E 




E, 


0° 


0° 




0° 


10° 


-13° 




-12° -5 


20° 


— 20°*9 




-21°-7 


30° 


-27°'l 




-27°-9 


40° 


-32°-2 




•— 31°-9 


43°30' 


— 32°-9 




-32°-5 


50° 


— 34°'3 




— 34°'6 


60° 


-35°-5 




-36°'4 


70° 


-37°-2 




-37°-6 


80° 


-38°-5 




— 38°*3 


90° 


-38°'5 




-38°-5 



* <p = angle of prism face with orthopinacoid 100. 

These figures indicate that for the first 40° from the ortho- 
pinacoid the extinction angle rises very rapidly while for faces 
near the clinopinacoid the variations are very slight. 

For the sake of comparison the theoretical values of the 
extinction angles indicated by the Michel-Levy formula* are 
listed under column E x . 

From the melt diopside crystallizes readily, usually in the 
form of radiating prismatic individuals intricately intergrown 
and overlapping. A characteristic microscopic feature is the 
presence of fine bubble-like inclusions or cavities throughout 
the crystallized mass. These cavities are either tubular in 
shape and parallel in a general way the prismatic elongations 
of the crystallites ; or they appear cutting across the sections 
in an irregular way not unlike the cavities in a section of 
worm-eaten wood. The cavities are probably due to the 
shrinkage accompanying the crystallization of diopside from 
the silicate melt. — Such air spaces in the crystals from CaCl 2 
fluxes were only rarely observed and are not characteristic 
of the same. 

The Intermediate Compositions. — In studying the prepara- 
tions of this series intermediate in composition between the 
compounds, the microscopic analysis has been directed along 
two principal lines : (1) To ascertain whether or not the prod- 
uct is homogeneous ; (2) to determine as accurately as pos- 
sible the optic properties of the one or more components in 
each preparation. Experience with both thermal and optical 
data has shown that in certain instances limits of homogeneity 
cannot be detected within one or two per cent optically and 
the optical determinations of the limits of solid solution in 
this series given below may easily be in error therefore one or 
two per cent. 

*Les Mineraux cles Boches, p. 11, 1888. In this formula the following 
values were used: V=29 n< 6, y=0°, ext. angle=38°-5 ; or //=8 0> 9, v=68-l ; 
u + v- 77°. 



to Calcium and Magnesium Metasilicates. 35 

When out of the melt of a readily crystallizing substance, 
minute quantities of a second substance crystallize, they are 
usually so completely hidden in the mass of the first crystals 
that the process of finding them microscopically is not unlike 
that of " finding the needle in a haystack," particularly when 
the optical properties of the two substances are closely similar. 
It has been found by experience that the best method for 
detecting inhomogeneity is to immerse the powdered material 
(finely divided by tapping the substance in a mortar) in a 
liquid of the refractive index of the predominating substance ; 
in this the minute particles of the second substance can be 
seen at a glance, if its refractive index be different from that 
of the first. For this purpose, a set of refractive liquids of 
indices ranging from 1 *450 to 1*790 has been used, the refrac- 
tive index of each successive liquid differing from the forego- 
ing by "005. The refractive indices of these liquids were 
determined directly on an Abbe-Pulfrich total refractometer and 
their constancy checked every three months at least.* The 
liquids are kept in small dropping bottles (30 cc capacity) 
with ground glass stopper and ground glass cap, and the 
refractive indices of the liquids change either very slightly or 
not at all in three months. 

In ascertaining the different optic properties of the members 
of this series, the following methods have been found most 
serviceable and seem best adapted to work of this character: 
Refractive indices were measured by use of refractive liquids 
after the immersion method of Schroeder van der Kolk. 
Wherever possible, especially on homogeneous preparations, 

* The following is the list of liquids used in the preparation of this set. 
(On an average the change in refractive index of a liquid is about *001 for 

in temperature.) 

Liquids used. 

Mixtures of chloroform and carbontetrachloride. 

Mixtures of turpentine and xylol. 

Mixtures of xylol and monochlorated benzene. 

Aethyliodide. 

Mixtures of cedar oil and clove oil. 

Monochlorated benzene. 

Mixtures of cedar oil and cloye oil. 

Mixture of aethylbromide and monochlorated benzene. 

Mixtures of clove oil and cinnamic aldehyde. 

Nitro-benzol. 

Benzene monobromated. 

Mixtures of clove oil and cinnamic aldehyde. 

Mixtures of benzene mono-chlorated and a- monochlorated 

naphthaline. 
1*640 to 1"655 Mixtures of a monochlorated naphthaline and a-mono- 

bromated naphthaline. 
1*660 to 1'740 Mixtures of a-monobromated naphthaline and methylene 

iodide. 
1/740 to 1*790 Mixture of methylene iodide and sulphur. 



a change of 3° G. 


Refractive Index, 


1-450 to 1-465 


1-470 to 1-495 


1-500 to 1-505 


1-510 


1-515 to 1-520 


1-525 


1-530 


1-535 


1-540 to 1-550 


1 555 


1-560 


1-565 to 1-615 


1-620 to 1-635 



36 



Allen, etc. — Diopside and its Relations 



the refractive indices were measured directly on the total refrac- 
tometer, experience having shown that with polished plates of 
the crystallized mass of substance the 7 and a limiting refrac- 
tive index lines can usually be distinguished if the reducing 
attachment be used. In such cases where the maximal and. 
minimal indices are determined from aggregates of crystals 
rather than from one crystal, it is of course not possible to 
determine also the refractive index /3. Optic axial angles 
were determined by use of the two-screw micrometer ocular 
and also by the modified universal stage* of Feclorow. Extinc- 
tion angle measurements were made with the aid of the bi- 
quartz wedge plate,f care being taken to select favorable 
sections in each case. Direct determinations of the birefrin- 
gence in the thin section or in flat crystal grains were accom- 
plished by measuring the thickness of the plate or grain with 
the micrometer screw of the microscope (model Fuess) and 
then ascertaining the path difference of the emerging waves 
by means of a Babinet compensator. By this method only a 
fair degree of accuracy can be obtained because of the 

Fig. 12. 



ou 










































70 










































T"P 


sV 


/-s- 














I — 
& — 


or 


U F 


!S r 


hr 


]T"~ 


-- 


Y-r.y 












^_ — rfX 


:mfs 


-X,_ 


WJ? 




















< 


* ( 


! r 


W- 


DlC 


V 


fcO 




T"> r 


\a/ 








' ' v^" 






1 5 




£> 






? 


V 






_/--£ 


Lp 


5 / 


T) 




\/ 


n 


M 


1^ 


if) 





















50 




























^ 


ti 


r V 


:_-J_- 




d-5. 


^ 










































4-0 


¥-- 


-ex 


...I 


D _b.. 


_w_ 


b__ 


--- 


----- 


























-t- 


\ 


Vb 
















































r 


■ 1 




h 


j 


^( 




















30 






































r 








D 


1 u 


H 


b 1 


JJ t 












- 1 




































.. 




P 


S 7 


TTl 






20 




J 


1 n 


V" 


r* 


\ » 


f r 


































10 


-H 




























X- 


o< 


P' 


M( 


1^ 


i0* 






































J 


j 













































I 20 50 40 50 ^0 70 80 ^0 100 



* This Journal, xxiv, 317-369, 1907. f This Journal, xxvi, 377-379, 1908. 



to Calcium and Magnesium Metasilioates. 37 

difficulty of determining the thickness of the plates accu- 
rately. This error was reduced so far as possible by taking 
the average of a number of measurements of thickness 
on the same section and by grinding the sections thicker 
than usual. 

The optical data prove that wollastonite can take up in solid 
solution about 17 per cent of diopside, while pseudo-wollaston- 
ite can absorb only about 4 per cent of diopside in solid solution ; 
that in diopside only a small amount of the calcium metasil- 
icate, not over 5 per cent, can enter into solid solution, while 
mixed crystals containing up to about 39 per cent of MgSi0 3 
in diopside can exist ; the maximal solid solution of diopside 
in /3-MgSi0 3 is not great and does not exceed 5 per cent. 
These relations were ascertained both by observing the limit of 
homogeneity of the preparations and by observing the changes 
in the different optic constants of the preparations. 

In fig. 12 the optic constants are represented graphically and 
the limits of solicl solution are indicated by the breaks in the 
curves. The optical data from which these curves were drawn 
are included in the following table (VIII). 

Although every effort was made to reduce the probable error 
of the values listed in this table, the very nature of the mate- 
rial precluded accuracy of a very high order. The refractive 
indices given in the table may be considered exact within 
±'003 ; the direct determinations of birefringence within 
±'003, especially when checked by refractive index determi- 
nation ; the optic axial angles are not all of the same order of 
accuracy, — that of pseudo-wollastonite being the least satisfac- 
tory, with /S-^IirSiOg next ; in general the probable error does 
not exceed ±1°; the extinction-angle determinations on 110 
probably vary less than 1°- from the true values. 

Table YIII as well as fig. 12 prove conclusively that diopside 
is a compound, and that the limits of solid solution for 
the different members of the series is that indicated above. 
Beginning with wollastonite, the refractive indices, the birefrin- 
gence and the optical angle of the pure compound increase with 
increasing admixture of MgSi() 3 up to about 8 per cent MgSi0 3 , 
after which the curves are practically horizontal, thus marking 
the limit of crystal miscibility with diopside. In the sections 
free diopside was observed first in the preparation containing 
10 per cent MgSi0 3 . — For pseudo-wollastonite the refractive 
indices, the birefringence and the optic axial angle increase 
slightly only up to about 2 per cent MgSi0 3 , and the exami- 
nation of the preparations proved that beyond this limit inho- 
mogeneity exists and diopside is present. — The limits of solid 
solution of CaSi0 3 in diopside were difficult to determine optic- 
ally with any degree of certainty. The data indicate only 
slight solid solution, probably not below 15 per cent MgSi0 3 , 
or not over 2 per cent of calcium metasilicate. In the prepa- 



3S Allen, etc. — Diopside and its Relations 

Table VIII. 



Composition 
% MgSi0 3 


a 

1-616 
1-609 

1-612 
1-620 

1-670 

1-664 
1-664 
1-662 
1-660 
1-654 

1-650 

1-645 
1-641 


1-628 

:: 

1-630 

1-671 

-" 

1-647 
1-648 


7 

1-632 
1-650 

1-654 
1-634 

1-694 

1-694 
1-692 
1-684 
1-684 
1-678 

1-660 

1-655 
1-663 


y— a* 


y-a 

•016 
•041 

•042 
•014 

•024 

•030 

•028 
•022 
•024 
•024 

•oYo 


Y-P 

•004 

•o"o"4 
•02~3 

•015 


7-/3' 
•012 

•o'fo 

•007 
•007 


Si 

o 

c 
+ 

+ 
+ 
+ 

+ 
+ 

+ 
+ 


Optical 

axial 

angle 

2 V 


Dis- 
per- 
sion 


Extinc- 
tion 
angles 

c : c' on 
110 





I 

5 

7-95 
7-95 

10 

10-74 

28 

28 

40 

44-5 

45 

46-12 

50 

55 

60 

65 

665 

68 

70 

72 

75 

80 

90 

95 

97 

98 
100 
100 


Wollastonite 

Pseudo-wollastonite 

a 

Wollastonite 

Pseudo-wollastonite 
Wollastonite 

a 

Diopside 

■ a 

\\ 

a 
a 
i i 
(< 
a 
a 
<( 
a 
a 

/?-MgSi0 3 

a-MgSiOs 


•014 
•043 

•043 
•014 

•030 

•030 

•027 
•024 
•022 
•023 

•023 


38 c 

7° 

10° 

10° 

38° 
41° 
7° 
40° 
40° 
41° 
57° 
58° 

58° 
59° 
58° 
57° 
56° 
54° 
53° 
53° 
52° 
51° 

52°" 

53°" 

53° 
53° 

60° 


8 
5 


5 



8 
2 
5 
5 
1 

5 

8 
5 
5 



6 
2 
5 
7 

5 

4 

5 

5 
R 


p > V 

p > V 

p > V 


-32 V -4 
-32° -5 
-32°-4 

-82°~8 

-31°'7 
-30° -2 
-28°-4 
-27°-9 

-27"°"-4 
-27° -4 

..27°"-4 























* y— a in this column was determined directly by measurements on plates in the thin 
section. 

rations between 40 and 45 per cent peculiar phenomena were 
observed and no satisfactory tests of homogeneity could be 
made. — Between diopside and pure magnesium metasilicate 
mixed crystals extend from diopside up to about 66 or 67 
per cent MgSi0 3 of the series. The appearance of all prepa- 
rations between pure diopside (46*12 per cent MgSi0 3 ) and 67 
per cent MgSi0 3 is that of diopside in the thin section so far 
as the optic properties are concerned. The refractive indices, 
the birefringence, the extinction angles and the optic axial 
angles all decrease gradually but noticeably, and the sections 
appear homogeneous under the microscope except for the 
minute air spaces. The crystallized melt changes noticeably 
moreover in appearance from the large, bright and glistening 



to Calcium and Magnesium Metasilicates. 39 

aggregates of cliopsicle to dull, lusterless, white granular masses 
of much finer grain and higher MgSi0 3 content. 

Beyond the 68 per cent the preparations appear inhomoge- 
neous, the /3-MgSi0 3 appearing in increasing amounts as its 
composition is approached. In the preparations ranging from 
68 to 90 per cent MgSi0 3 in composition, /3-MgSi0 3 appears 
almost without exception intergrown with the diopside and 
usually occupies the center of the large diopside sections. The 
/3-MgSiO s is invariably twinned polysynthetically after 100 and 
on sections approximately normal to the prism axis, the pris- 
matic cleavage lines can be seen cutting across both the diopside 
and the /3-MgSi0 3 , thus indicating, the close crystallographic 
symmetry of the two compounds. In such sections the /3-MgSi0 3 
is characterized by its weak birefringence and by the position 
of its optic axial plate parallel with the twinning lamellae, while 
for the enclosing diopside substance the optic axial plane is at 
right angles to the twinning lamellae. — The limit of crystal 
miscibility of diopside in /3-MgSi0 3 is above 98 per cent MgSi0 3 , 
since preparations of that composition are clearly inhomoge- 
neons. The optic properties of the /3-MgSi0 3 change slightly 
between 98 and 100 per cent Mg$i0 3 , but only enough to indi- 
cate very slight solid solution, probably not over 2 per cent 
of diopside. 

Etch figures. — Proof of the fact of solid solution for composi- 
tions ranging from pure diopside to about 67 per cent MgSi0 3 
was also gained by etching the crystals with hydrofluoric acid. 

At the present time isomorphism is a much discussed subject 
and its final definition has not yet been agreed upon. Emphasis 
has been placed on similarity of the crystal form of the two end 
members, on analogous chemical composition, on complete misci- 
bility, and on the fact that for some of the physical properties, 
as specific volumes, the properties of the intermediate mixtures 
are additive functions of those of the end members. This 
last assumption of "Retgers has been questioned recently, * while 
the qualification of analogous chemical composition has long 
been considered unnecessary by certain investigators. 

Briefly stated, the tendency seems to exist for crystallizing 
substances to absorb, during the process of crystallization, large 
or small amounts of other material. The more closely similar 
the absorbed material is to the absorbing crystal in crystal 
structure, dimensions and tendencies, the greater the amount 
in general which can be thus taken up. In certain instances, 
the properties of two substances are so similar and their molec- 
ular volumes so nearly equal, that the solid solution or crystal 
miscibility extends from the one compound without break to 
the second, and the physical properties vary continuously 
throughout the series. In such a case of complete miscibility 
no doubt can exist as to the isomorphic relations of the two end 
*B. Gossner, Zeitschr. Kryst., xliv, 417-519, 1908. 



40 Allen, etc. — Diopside and its delations 

members. But in case the crystal miscibility is incomplete or 
limited the term isomorphism has not yet been defined with 
adequate precision nor the criteria therefor developed with 
sufficient sharpness to permit one to state in every actual case 
whether or not isomorphism does exist. It seems proper, how- 
ever, to speak of the two compounds in the first case as com- 
pletely isomorphous, while in other instances the isomorphism 
is incomplete or limited. In the case of incomplete isomorphism, 
criteria such as crystallography similarity, additive character of 
certain physical properties in the intermediate mixtures, chem- 
ical analogy are then relied on to prove isomorphism. 

In the present series, the diopside and /3-MgSi0 3 are of the 
same crystal system and somewhat similar in crystallographic 
properties. Limited miscibility has been shown to exist and 
the physical properties prove that the two are incompletely 
isomorphous. This being the case, it is reasonable to suppose 
that their internal crystal structure is similar, that the distri- 
bution of the effective crystallographic forces is analogous. 
In the intermediate mixtures, therefore, the distribution of the 
forces is represented, approximately at least, by the resultants 
of the crystallographic forces of the end members in their proper 
intensities. One of the best methods for studying the distribu- 
tion and relative intensity of crystallographic forces is by means 
of etch figures. On the above assumption, the etch figures of 
the intermediate members should be intermediate in character 
between those of the end members, diopside and /3-MgSi0 3 .* 

After considerable experimentation on the conditions best 
suited to produce favorable results, both with respect to the 
etch figures and the handling and photographing of the exceed- 
ingly small crystals, this statement has been substantiated. The 
etch figures on 110 were produced by immersing cleavage 
pieces from the different preparations in hot commercial hydro- 
fluoric acid (heated over a steam bath in a platinum crucible) 
for 40 seconds and then stopping the reaction by plunging the 
crystal into cold water. The crystal was then mounted on the 
condenser lens attachment of the universal stagef and examined 
in strong reflected arc light and turned until the proper cleavage 
face, 110, was normal to the axis of the microscope. The etch 
figures on 110 thus obtained were not of equal size or develop- 
ment for the different members of the series- The largest and 
best developed etch figures are those of diopside (Plate I, 
photomicrographs a and h), while the least favorable are those 
of /3-MgSi0 3 , which are exceedingly difficult to obtain under 

*The value of etch, figures in determining isomorphism has been strongly 
advocated by Eetgers (Zeitschr. f. Phys.-Chem., xvi, 35, 1895) and notwith- 
standing the objections which have been raised to this criterion, it does 
apply in certain series. In the present case of limited isomorphism the etch 
figures sustain the contention of Eetgers. 

f This Journal, xxiv, p. 342, and fig. 7, p. 332, 1907. 



to Calcium and Magnesium Metasilicates. 41 

any conditions and are excessively small. The photomicro- 
graphs (Plate I), a-f\ illustrate the changes in the shape of the 
etch figures thus produced on 110 in preparations ranging from 
diopside to pure /3-MgSiO a . The etch figures on diopside (a 
and o) are long spindle-shaped pits with an upper blunt termina- 
tion. On an average their length is four times the width and 
the angle a between the two sides at their point of junction at 
the lower extremity is about 18°. These etch pits are similar 
in every detail to those on 110 of natural diopside from Ala,* 
Piedmont. Etch pits on the face 010 were also observed and 
likewise resembled those on 010 of the natural mineral. — The 
etch pits on 110 of the preparation 50 per cent MgSi0 3 (c) are 
very similar to the figures on diopside, the angle a between the 
two sides at the lower extremity being slightly greater perhaps. 
On crystals of the composition 55 per cent MgSi0 3 (d) the etch 
pits are noticeably wider and the angle a has increased to about 
35°. This angle a gradually increases until at the limit of solid 
solution it measures approximately 50° {f). The relation of the 
length of the etch pits to their width changes from about 4 : 1 in 
diopside to 3 : 1 in the 55 per cent and about 5 : 3 at the limit as 
represented by the etch figures of the 75 per cent preparation, 
which were chosen in place of those of the 67 per cent because 
of the sharper definition of the particular figures photographed. 
The transitional changes in the shape of the upper portion of 
the etch figures are also characteristic. The etch figures on 
the cleavage faces of /3-MgSi0 3 (<?) are exceedingly small and 
triangular in shape ; the angle a measures 55°-60° and the rela- 
tion of length to width is about 3 : 2. Other details of the etch 
figures appear in Plate I and the evidence from all view- 
points tends to strengthen the assertion, that in this case of 
limited or incomplete isomorphism the character of the etch 
figures does change continuously with increasing MgSi0 3 , and 
in the direction of the type of the etch figures of pure /3-MgSi0 3 . 

It is of interest to note that the /3-MgSi0 3 , the low tempera- 
ture form, takes up very little if any diopside in solid solution. 
This may be due to the fact that the high temperature-, a-form, 
is orthorhombic and therefore would have less tendency to take 
up the diopside molecule. 

On the calcium side of diopside the crystal miscibility in the 
series is very slight, diopside taking up only small amounts 
of the calcium metasilicate. Wollastonite, on the other hand, 
can absorb up to 17 per cent of diopside and still remain homo- 
geneous, whereas pseudo-wollastonite takes only 1 per cent at 
most of diopside in solid solution. ]N~o satisfactory explanation 
has been found to account for these differences in the behavior 
of the different compounds of this series. 

* Compare E. A. Daly, Proc. Amer. Acad, of Arts and Sciences, xxxiv, 
5373-428; pi. iv, No. 18, 1899. 



4:2 Allen, etc. — Diopside and its ^Relations 

In Part I of this paper, the fact of eutectic mixtures and 
its bearing on the present problem are shown to be of prime 
importance. In the case of alloys eutectic textures are 
definitely recognized and it is natural to expect such textures 
in silicate melts. In the latter, however, power and rapidity 
of crystallization, combined with viscosity, frequent absence of 
stable equilibrium, and other factors, tend usually to veil 
effectively such textures which might otherwise develop. 
Indications of probable eutectic textures were occasionally 
recorded in this series, but as a rule the crystallization appar- 
ently takes place so rapidly with strong undercooling that 
normal, theoretical conditions of equilibrium do not exist. 

It is also of interest to note that throughout this series 
the melts of the pure compounds are, as a rule, of coarser 
grain than the intermediate compositions and of more vitreous 
luster. The intermediate mixtures frequently resemble porce- 
lain in appearance. This change in aspect undoubtedly results 
from the decrease in granularity and lack of continuity of the 
single crystallites. 

The inversion of the a- and /3-MgSi0 3 . — With pure MgSiO s - 
melts the thermal data show practically no heat effect at the 
temperature of inversion of the fi- into the a-form and vice 
versa, and only after the admixture of several per cent of 
diopside does the thermal effect produced by the inversion 
appear. The quenching experiments of Dr. A. L. Day, how- 
ever, proved definitely that the a-MgSi0 3 did exist, and 
attempts were then made to fix the temperature of inversion 
by use of a specially constructed thermal microscope.* (Fig. 
13, a and b.) 

* Constructed in the workshop of the Geophysical Laboratory after plans 
by Dr. Arthur L. Day and the writer. The details of construction of the 
electric resistance furnace are given in fig. 136, two important features of 
which were suggested by Dr. Day, namely, the enclosing of the whole in 
a suitable water jacket and the splitting of the thermoelement wires to serve 
as a support for the preparations. By this latter device the purity and 
temperature of the preparations even at high temperatures is insured. The 
microscope is fitted with revolvable nicols ; at the base of the furnace there 
is a thin metal slide, A 3 , by means of which part or all of the field can be 
shaded and the characteristics of the light emitted and transmitted by the 
body studied with respect to effects of polarization. A plate at high tem- 
peratures may become often self-luminous and it is then necessary to 
adopt special devices to detect transmitted polarized light and with this end 
in view the optical system of the microscope has been arranged. With it, 
also, the character of the emitted light alone can be examined with respect 
to polarization effects if such exist. At high temperatures the white heat of 
the furnace tends to veil the interference phenomena unless the transmitted 
light be of greater intensity, and this condition has been met by using an 
electric arc as source of light. With this furnace the birefringence of 
quartz has been measured up to 1300° and it is proposed to study the optical 
changes in several minerals at different temperatures in this way. — Tempera- 
ture readings are made either roughly on a direct reading Siemens and 
Halske voltmeter or accurately by use of the potentiometer. — The water 
jacketing of the furnace permits its use on any microscope in which the 
distance between the stage and the objective can be made great enough, the 
optical system remaining thereby unchanged. 



to Calcium and Magnesium Metasllicates. 43 

Fig. 13, a. 




Fig. 13, 6. 




41 Allen, etc. — Blopside and its Relations 

Figs. 13, a and b. — Microscope equipped with electric resistance furnace. 
With the exception of the base and the upper tube, which were taken from 
a Fuess universal stage microscope, this thermal microscope was constructed 
in the workshop of the Geophysical Laboratory. The microscope is fitted 
with revolvable nicols and a low power 3-inch objective. The furnace rests 
directly on the stage of the microscope and can be revolved through small 
angles. Its different parts are shown in cress section in fig. 13& (one-half 
actual size). The water jacket consists of three parts ; W 2 , base ; D, side 
cylinder, and ¥,, cap. Each of these parts is complete in itself and by 
means of rubber tubing (fig. 13a) the water is made to pass through W 2 , then 
D and finally W } . At G in both Wi and W 2 glass plates are introduced 
and allow light rays to be transmitted without permitting the heat to reach 
the objective and lower condenser system. The circulating water is suffi- 
cient to keep these plates cool. In the first experiments air bubbles from 
the water collected between the plates and seriously disturbed the clearness 
of vision. This difficulty was overcome by means of rubber plungers fast- 
ened to Ai and A 2 (fig. 13) which could be passed back and forth in front of 
the plates and the bubbles brushed aside. By means of the rod A 3 with its 
attached brass plate, transmitted light from any part of the field can be shut 
off and the effects of emitted light alone studied. The furnace itself con- 
sists of a tube K (fig. 136) 7"5 cm long, 4 , 5 cm outside diameter and l cm inside 
diameter, wound on the inside with fine platinum wire - 35 mm diameter. The 
thermoelement wires are supported by the porcelain tube T, which rests 
directly on the asbestos paper covering the upper plate of W 2 . The thermo- 
element wires are introduced into the furnace at M (fig. 13a) and the furnace 
wires on the opposite side of the microscope. The sides of the furnace, 
F, are surrounded with magnesia powder and the ends capped with asbestos 
paper, to prevent loss of heat from radiation so far as possible. 

Better results were obtained by another method, in which 
single 5 water-clcar crystals of /3-MgSi0 3 (about *2 X '2 X l mm ) were 
mounted in cedar oil on the universal stage and turned until the 
clinopinacoid was normal to the line of vision and the twinning 
planes appeared as sharp lines. After photographing in this 
position (magnification 100 diameters) the crystal was placed 
in a specially prepared platinum basket and heated in an 
electric resistance furnace to a specified temperature, either 
above or below that of the /3-MgSi0 3 inversion. After cool- 
ing, the crystal was again photographed under precisely the 
same conditions as before heating. The study of a long series 
of negatives prepared in this way has brought out several 
interesting points : In the inversion of a single crystal of 
/3-MgSi0 3 to the a-form, no great volume change is involved 
nor even a great redistribution of the molecules. This is 
evident from the fact that after reversion from the a- to the 
/6-MgSi0 3 the original crystal is intact and its faces still fairly 
sharp. Twinning planes are still present though usually in 
different positions, each lamella extending the entire length of 
the costal as before heating. On such paramorphic change, 
inversion into one form and reversion to the original, it might 
be expected that, as in crystal aggregates formed by precipita- 
tion, many crystal nuclei would be formed,* and that on 
reversion each one of the new nuclei would produce at least 
one separate individual of the original form, with the result 

*This actually happens on the inversion of wollastonite into pseudo- 
wollastonite. This Journal, xxi, 107, 1000. 



to Calcium and Magnesium Metasilicatex. 45 

that instead of a single crystal or a regularly twinned crystal, 
the aggregate of irregularly oriented individuals would result. 
In case, however, the molecular redistribution was slight, the 
inversion might proceed in regular fashion throughout the 
entire crystal and the effect of inversion and reversion be 
chiefly one of shifting of the ever present twinning lamellae. 
And this is the exact state of change in the /3-.MgSi0 3 crystals. 
Rarely the subdivision of a crystal into several irregularly 
bounded parts was noted and usually only the shifting of the 
lamellae. It was also of interest to observe that occasionally 
a shifting of the twinning lamellae took place in crystals heated 
to temperatures slightly below the inversion point. Because 
of this property, no decisive determinations of the inversion 
point could be made and recourse was taken to sudden chilling 
experiments — the preparations being first melted, the tempera- 
ture then lowered and kept at a specified point for one hour, 
after which the preparation was dropped into cold water and 
chilled almost instantly. The a-MgSi0 3 thus obtained clearly 
showed the effect of incipient change even under these condi- 
tions, the major part of the powder being full of minute 
dustlike particles or cavities and as a result was semi- or sub- 
transparent, and only now and then were clear portions of the 
a-form observed. Whenever the a-MgSi0 3 was held at tem- 
peratures slightly below the inversion point and then quenched, 
the entire preparation consisted essentially of the twinned 
/3-form alone in clear transparent individuals, the dusty effect 
as well as the a-MgSi0 3 aggregates having practically disap- 
peared except for an occasional clear crystal of the same. 

Summary. 

1. The end members of the system CaSi0 3 -MgSi0 3 both 
exhibit enantiotropy. The inversion point in the former is 
about 1190°. The a-form, pseudo-woUastonite, is unknown 
in nature. The /3-form is the mineral wollastonite. The 
/3-form of magnesium silicate is the magnesian pyroxene occur- 
ring in meteorites and in intergrowths with enstatite and has 
recently been called clino-enstatite.* At about 1365° it is 
transformed into an orthorhombic form quite distinct from 
enstatite and unknown in nature. 

2. Only one stable compound appears, viz., CaSiG 3 .MgSi0 3 , 
identical with diopside. It melts at 1380° and has a specific 
gravity of 3*275. It was obtained in well-formed, measurable 
crystals extremely pure, when crystallized from molten calcium 
chloride. 

3. A eutectic occurs between diopside and pseudo-wollas- 
tonite at the composition 60 per cent diopside : 40 per cent 
calcium silicate. It melts at 1348°. A second eutectic occurs 
at about 63 per cent ]\IgSi0 3 : 32 per cent CaSi0 3 . It is com- 

* W. Wall], Die Eustatit-augite, Tscherinak's Mitth., xxvi, 1-131, 1907. 



46 Allen, etc. — Diopside and its Relations 

posed of about 95*5 per cent of a mix-crystal containing about 
62*5 per cent of diopside, 37*5 per cent magnesium silicate, 
and 4'5 per cent a-MgSi0 3 . Its melting temperature is 1375°. 
Microscopically eutectic textures were observed rarely if at all. 

4. Six solid solutions appear in this system. Only two of 
them contain more than three or four per cent of the lesser 
component, and only these will be mentioned here. 

a. ^-calcium silicate (wollastonite) forms a saturated solution 

of wollastonite containing about ll per cent diopside (8 per cent 

MgSi0 3 ) : 83 per cent CaSi0 3 , when crystallization takes place 

in the neighborhood of 1050°, i. e., wollastonite is capable of 

17 
dissolving about — =20 per cent of its own weight. This series 

83 

of solutions is interesting from the fact that the inversion point 
of pure calcium silicate (1190°) appears to be raised by the addU 
tion of MgSi0 3 , up to 100° in the most concentrated solutions. 
This is probably largely, if not wholly, an apparent rise in the 
inversion point due to viscosity, for, as is well known, an inver- 
sion point should be raised only when the concentration of the 
solution below the point is greater than that above, while here 
there is a rise in the weaker solutions which suffer no change 
in concentration when they invert. Again, the concentration of 
solutions just below the inversion point cannot be determined 
with accuracy on account of the difficulty of establishing an 
equilibrium in solid silicate solutions. 

b. Diopside dissolves about 60 per cent of its own weight, 
forming a solution which contains 66"5 per cent MgSiO„ : 33'5 per 
cent CaSi0 3 . This saturated solution is very similar to diopside 
in all its properties. Its melting point is only 3° lower (i. e., the 
maximum heat absorption falls there, the melting interval is 
unknown). The specific gravity changes very little ; the optic 
data show slight but noticeable changes : the refractive indices, 
the birefringence, the optic axial angle and the extinction angles 
all falling continuously with the addition of MgSi0 3 from diopside 
up to the limit of solid solution at 66-5 per cent MgSi0 3 . 

5. In the series of limited solid solution between diopside 
and clino-enstatite, the effect of the addition of MgSi0 3 to 
diopside is, furthermore, clearly shown by etch figures on the 
prismatic cleavage faces. On passing from diopside to the 
limit of solid solution at about 66*5 per cent MgSi0 3 , the 
shape of the etch pits changes gradually, their character, on 
preparations of intermediate composition, being intermediate 
between those of the two compounds, diopside and pure 
/3-MgSi0 3 , thus proving that actual solid solution does exist in 
the series, and that the effects of the end members are felt 
crystallographically in the solid solutions of the same. — For 
the observation, under the microscope, of changes which take 
place in substances at high temperatures, a special micro- 



to Calcium and Magnesium JMetasilicates. 47 

scope, fitted with electric resistance furnace, fig. 13a, lias been 
constructed and found useful in the study of these etch pits. 

6. The specific-volume curve consists of three well-defined 
branches, the first of which is the locus of the volumes of 
mechanical mixtures of pseuclo.-wollastonite (a-CaSi0 3 ) and 
diopside ; the second, that of the solid solutions of magnesian 
pyroxene (/3-MgSi0 3 ) in diopside ; and the third the locus of 
the volumes of mixtures of saturated mix-crystals just men- 
tioned, and the free magnesian pyroxene. The volume of the 
solid solutions is greater than the sum of the constituent 
volumes. There is a sharp minimum on the curve at the 
composition of diopside CaSi0 3 .MgSi0 3 . On account of the 
presence of minute bubbles in the crystals and the compara- 
tively small difference between the specific gravity of diopside 
and that of the magnesian pyroxene, the critical points on the 
curve are several per cent in error. 

7. A method for the more accurate determination of the 
composition eutectics is described ; also a method for the 
approximate location of inversion points in inert substances. 
The accidental variations between different determinations of 
the melting point of a sharp melting silicate seldom amount to 
1° up to 1500°. This is the accuracy available for compara- 
tive measurements. The absolute accuracy of a determina- 
tion is less than this on account of the present limitations of 
the absolute scale. 

The authors wish to express to Dr. Arthur L. Day their 
hearty thanks for valuable assistance in connection with the 
study of the a-magnesium silicate. 

Geophysical Laboratory, Carnegie Institution of Washington. 
Washington, D. C, July 10, 1908. 



48 Gilbert — The California Earthquake. 



Art. II.— The California Earthquake of 1V06 ;* by G. K. 

Gilbert. 

Three days after the California earthquake of April 18, 
1906, Governor Pardee appointed a commission for its scien- 
tific investigation. ~No funds were at his disposal to defray the 
expenses, but provision was made later by the Carnegie Insti- 
tution, and the Institution is publishing the reports. Volume 
I, in two parts with atlas, has recently appeared, and a second 
volume is to follow. 

Volume 1 is by Andrew C. Lawson, chairman of the com- 
mission, and includes contributions from a large number of 
collaborators. After an introductory account; of the geology 
and morphology of the Coast Ranges, it treats at length of the 
physiographic features and physical changes associated with 
the earthquake, of the distribution of intensity, and of the 
directions of vibratory motion. The marine phenomena, the 
composition of the main shock, the sequence of after shocks, 
and various minor topics are presented, and account is given of 
earlier severe earthquakes in the same region. 

The earthquake was of the tectonic class, and was occasioned 
by a slipping on the plane of an old fault. The fault outcrops 
at the surface, and there was a visible displacement of consid- 
erable amount. The line of outcrop trends NW.-SE., and 
the fault plane is vertical. There was, however, very little 
vertical displacement, the differential movement being almost 
wholly horizontal. The country adjacent to the fault on the 
SW. side moved bodily toward the NW., and the country 
on the NE. side moved toward the SE. The changes did 
not tend to increase the height of a mountain or the depth of a 
valley but merely to distort the land horizontally. The amount 
of displacement was measured in two ways, (1) by observation 
of the dislocation of roads, fences, etc , traversed by the fault, 
(2) by the remeasurement of a net of triangulation previously 
made by the Coast Survey. Fences and roads were usually 
offset from 8 to 15 feet, and the results from triangulation 
showed relative dislocation of about the same amount for 

* The California Earthquake of April 18, 1906. Report of the State Earth- 
quake Investigation Commission. In two volumes and atlas. By Andrew 
C. Lawson, chairman, in collaboration with G. K. Gilbert, H. F. Eeid, 
J. C. Branner, H. W. Fairbanks, H. 0. Wood, J. F. Hayford and A. L. 
Baldwin. F. Omori, A. O. Leuschner, George Davidson, F. E. Matthes, E. 
Anderson, G. D. Louderback, R. S. Holway, A. S. Eakle, R. Crandall. G. F. 
Hoffman, G. A. Warring, E. Hughes, F. J. Rogers, A. Baird, and many others. 
Vol. I, pp. xviii + 451. 146 pis. Atlas, 25 maps, 15 pis. seismograms. 
Washington, D. C, 1908. (Published by the Carnegie Institution of Wash- 
ington.) 



Gilbert — The California Earthquake. 49 

points near the fault line. For points at greater distance the 
changes were less. The discussion of the data, bj J. F. Hay- 
ford and A. L. Baldwin, led to the conclusion that the absolute 
movement was greater west of the fault than east of it, and in 
both directions diminished with distance from the fault, the dimi- 
nution being most rapid in the immediate vicinity of the fault. 

This earthquake is practically unique, among the small 
group that have been broadly studied, in that the stress couple 
to which the fault may be referred lay in the horizontal plane. 
The main associated distortions were distortions in ground 
plan, with little vertical complication. They were, therefore, 
exceptionally adapted for measurement by the method of 
triangulation, and the results actually obtained are more syste- 
matic than any previous results of the same character. It is, 
therefore, peculiarly unfortunate that they were qualified by 
a lack of chronologic unity in the trigonometric surveys 
preceding the fault, which were strung along through several 
decades. This fact made it impossible to discriminate between 
deformation at the time of rupture and progressive deforma- 
tion during accumulation of strain before rupture ; and if 
progressive deformation took place before rupture, the pre- 
cision of the adjusted triangulation was thereby impaired. 
Nevertheless the results invite the careful attention of geophys- 
icists. To the reviewer the distribution of dislocation, and 
especially the existence close to the fault, on each side, of a 
belt of maximum distortion, seems clearly not that which 
would obtain if the fault passed completely through a solid 
crust to a liquid substratum. And it appears also that, on the 
assumption of continuous solidity from the surface downward, 
the geodetic results might yield to adequate analytic treatment 
a conception of the order of magnitude of the vertical distance 
to which the fault penetrated. 

The surface outcrop of the fault was definitely traced from 
San Juan to Point Arena, a distance of 190 miles. At Point 
Arena it passes under the sea, and there is doubt as to its 
further course. A fault made at the same time on a more 
northerly part of the coast may be its continuation, after 
inflection, or may be on an independent line; but in either 
case the total length of dislocation was about 270 miles. 

At all points the fault follows a peculiar topographic feature 
to which the name San Andreas rift was given ; but the rift 
is more extensive than the fault of 1906, having been traced to 
the Salton basin, several hundred miles southeast of San Juan. 
In its larger expression the rift is a trough, a trough coin- 
ciding in general trend with the Coast Eanges, but crossing 
various mountain ridges obliquely, or even following their crests. 

Am. Jour. Sci. — Fourth Series, Vol. XXVII, No. 157. — January, 1909. 
4 



50 Gilbert — The California Earthquake. 

In detail it comprises many small ridges and hollows, approxi- 
mately parallel but otherwise irregularly disposed, and evi- 
dently caused by splintery dislocation. Streams zigzag more 
or less about the ridges, and the hollows contain many small 
ponds and marshes. There are reports of long cracks which 
appeared in different parts of the rift in connection with various 
earthquakes of the last century, and it is inferred that each of 
these cracks was the surface expression of a fault-slip similar 
to that of 1906. It is further inferred that the rift as a whole 
marks the outcrop of a long fault or fault zone, separating two 
crustal tracts which are slowly moving past one another, with 
gradual accumulation of strain and stress, and occasional relief 
by local slipping when the stress at some point overpowers 
the adhesion on the fault plane. The physiography of the rift 
is illustrated by numerous excellent photographs, and by a local 
contour map by F. E. Matthes. Although the rift has been 
mentioned in various writings of earlier date, its description in 
this volume practically adds a type of surface configuration to 
physiographic science. 

In the discussion of the intensity of the shock, a distinction 
is recognized between the elastic wave propagated from the 
origin through the crust, with gradually diminishing magnitude, 
and the phenomena of emergence, conditioned by the nature 
of the surface formation. The intensity observed at the surface, 
and expressed chiefly by damage to buildings and other struc- 
tures, is called "apparent intensity," and this only is mapped. 
The general map shows a long narrow belt of high intensity, fol- 
lowing the fault, with peninsulas and outlying islands where 
destructive effect was enhanced by the presence of incoherent 
formations; but this elongation is less characteristic of the 
lines limiting the areas of low intensity. The outer line, 
touching the most remote points of sensible tremor, traverses 
southern Oregon, central Nevada and southern California. 
In view of the ideas recently advanced by W. H. Hobbs, 
there is a careful review^ of the relation of local intensity to 
the known major faults of the region, about forty in number. 
In three cases it was thought possible that some portion of 
the movement of dislocation was diverted from the main (San 
Andreas) fault to the planes of intersecting faults. A special 
intensity map of San Francisco, by H. O. ^W\)od, shows with 
great detail the grades of violence; and its comparison with a 
geologic map brings out forcibly the intimate relation between 
effective intensity and the underlying formation. 

The subject is further elucidated by the report of an experi- 
mental study by F. J. Rogers. By mechanical arrangements 
similar to those employed by the Japanese commission in inves- 
tigating the principles of earthquake-proof construction, bar- 



Gilbert — The California Earthquake. 51 

monic horizontal motion was given to an open box containing 
some loose material such as sand. A block resting on and 
anchored to the upper part of the sand, so as to share its motion, 
was found not to have harmonic motion, but motion of a dis- 
tinct type which varied with the conditions of the experiment. 
Under certain conditions the amplitude of its motion was 
greater than that of the motion of the box, and its maximum 
acceleration — the factor corresponding to earthquake inten- 
sity — very much greater. These novel experiments are not 
only valuable in their immediate results, but of signal import- 
ance as indicating a line of study which should develop a com- 
plete theory of the phenomena of the emergence of earthquake 
waves. 

The marine phenomena were in accord with the terrestial 
in that they indicated no bodily movements of the ground 
except in a horizontal sense. Vessels at sea experienced a 
shock ; there were boilings of water near the shore ; a small 
seiche was started in San Francisco bay ; a wave several feet 
high washed the east shore of Tomales bay, a narrow sheet of 
water traversed by the fault ; but there was no great sea wave 
such as accompany vertical dislocations of the ocean bed. 

The main shock, which was of about one minute duration, 
was reported by many observers as consisting of two parts, or 
having two maxima, but by others as continuous. Consider- 
ing the improbability that movement was synchronous and 
similar over the entire plane of rupture, it is to be assumed 
that the vibration had different characters at different places, 
but the observations are not discussed with reference to geo- 
graphic distribution. There are many records of preceding 
or accompanying sounds, all of low pitch. The after shocks 
were of normal character, diminishing with time in frequency 
and average strength, and continuing for at least ten months. 
The report enumerates more than 100 in the Urst 24 hours ; 
about 300 in the first month ; and for succeeding months, 
71, 21,44, 28, 14, 11, 13, 15, 21, 2, 3, 1, 2. The record is 
recognized as fragmentary, and the actual number of sensible 
shocks was probably much larger. There was somewhat volu- 
minous testimony to the occurrence of visible undulations of 
the surface of the ground, the speed of which was much slower 
than that of the elastic waves in rock. 

Cracks opened in many places near the fault ; from several of 
these were large temporary discharges of water or of water and 
sand; the circulation of underground water was seriously and per- 
manently deranged, springs being destroyed, created or changed 
in volume ; landslides and earthflows were precipitated in great 
number. Alluvial lands slumped toward stream channels, and 
soft ground was in some localities left with a wavy surface. 



52 Gilbert — The California Earthquake. 

The volume closes without discussing the subject of future 
earthquakes in the San Francisco region, but furnishes material 
pertinent to that discussion by publishing accounts of the earth- 
quakes of 1868, 1865 and 1857. The fault in 1857 was on the 
southern part of the San Andreas rift, and the fault in 1865 
may have been on the same rift near the southern end of the 
fault of 1906. The fault in 1 868 was on a parallel rift east of the 
southern arm of San Francisco bay. In each case the distri- 
bution of intensity in San Francisco was substantially the same 
as in 1906, the character of the ground having more influence 
than the direction of the origin. 

The second and closing volume of the report will be by H. 
F. "Reid, and will treat of the theory of the seismogram. 



T. D. A. Cocker ell — Descriptions of Tertiary Insects. 53 



Art. III. — Descriptions of Tertiary Insects ; by T. D. A. 
Cockerell. Part V. 

Some New Diptera. 
Lasiosoma mirandula sp. nov. (Mycetophilidse : Sciophilinse). 

Expanse 16 mm ; length of body about 10 mm , of wing 7 mm ; 
abdomen reddish, with the hind margins of the segments rather 
broadly blackened, its whole surface with minute appressed 
hairs ; legs pale brown, the femora minutely hairy ; wings 
wonderfully preserved, strongly rufescent, with a purple or 
pink tinge, nervures darkened. By the dusky wings and the 
short prostrate hairs on abdomen it agrees with the living I. 
pallipes (Say), but it is considerably larger than Say's species. 
The venational features are as follows : the costal thickening 
extends a short distance beyond the tip of the ultimate branch 
of the radius (third vein) ; subcosta, much as in the fossil 
Sciophila hyatti Scudd., does not distinctly reach the costa, 
but is evanescent after the little vein which passes from it to 
the radius, meeting the beginning of the inferior branch of the 
latter ; small vein from costa to subcosta distinct, oblique, 
forming an angle of 45° with each, and placed a moderate dis- 
tance (about f mm ) from base of wing ; branches of radius as in 
Sciophila, with the same little cell formed in the fork by 
R 2+3 running up to join E, a : media, as usual, weaker than 
radius, and forking almost immediately after leaving the radio- 
medial cross-vein, so that the stem before the fork is shorter 
than the cross-vein, although the latter is short and not very 
far from vertical ; cubitus forking not far from base as in 
Sciara / anal with its apical portion wanting, as in Ceroplatus. 
In Williston's tables (N. A. Diptera, 3d ed.) this runs exactly 
to lasiosoma, and seems to accord well with that genus.* The 
living American species are northern. 

Hah. — Florissant, Colorado, in the Miocene shales, Sta. 14 
{S. A. Rohwer, 1907). 

Holotype in Peabody Museum (Yale). 

Tetragoneara peritida sp. nov. (Mycetophilidse : Sciophilinss). 

Length 4 mm ; wing 3-f, antenna l| mm ; dorsum of head, tho- 
rax and abdomen dark brown, the rest of the body pallid ; 
legs long and slender; wings hyaline, with brown nervures. 

*It differs, however, from Handlirsch's figure (Foss. Ins., pi. vi) of Lasio- 
soma in the shorter subcosta, not ending on costa, and also in the more ver- 
tical radio-medial cross-vein. Handlirsch's figure does not show any cell in 
the forks of the radius, so it may not be of this genus. 



54 T. D. A. Cocl.erell — Descriptions of Tertiary Insects. 

Antennae with about 16 joints, the middle joints much broader 
than long (about 102 /jl broad, and 76 long) ; scntellar region 
with two very long stout bristles, 459 \x long, the other bristles 
of thorax much shorter ; thorax very strongly convex above ; 
legs and abdomen with extremely fine hairs ; hind tibia3 with 
a row of short black bristles, about 93 /jl long. Yenation : sub- 
costa reaching to about the middle of the wing, that is to say, 
not abbreviated, and ending in a very broad Y, the upper 
branch extending to the costa, and the lower to the radius ; 
radius normal, its lower branch leaving it a short distance 
before the termination of the subcosta, and emitting, shortly 
after that termination, the oblique cross-vein to R a which 
really represents R 2+3 ; R 4+5 normal, ending very near apex of 
wing ; oblique radio-medial cross-nervure very long, about 
three times as long as the nearly vertical first section of lower 
branch of radius ; media with a long fork ; cubitus not dis- 
tinctly preserved (it ought to be forked). 

Hab. — A few miles north of Rifle, Colorado, in rocks of 
Eocene age, probably of the Green River Group, but possibly 
Wasatch. Type in collection of Dr. S. M. Bradbury. This 
cannot be identified with any of Scudder's fossil Mycetophi- 
lidee. Professor O. A. Johannsen very kindly examined my 
sketch and notes, and suggested the generic reference. He 
writes as follows : 

" The subcosta in most of the species thus far described is 
short, and ends either free, or in R ; in one species described 
by Walker it ends in the costa ; in one species of Meunier it is 
long, though ending in R. As only about sixteen species 
(including fossils) are known of this genus, and these quite 
rare, its limitations are difficult to define. I should be inclined 
to call your fossil Tetragoneura, though acknowledging it as 
somewhat aberrant" (litt. Aug. 19, 1908). The known fossil 
species of Tetragoneura number seven, all from Baltic amber, 
described by Meunier. As the amber is of Oligocene age, the 
species now described is the oldest known. 

Alepidophora g. nov. (Bombyliidse). 

A genus with elongate, subcylindrical abdomen, looking not 
unlike a bee. In Williston's tables (N. Am. Dipt., 3d ed., p. 216) 
it runs to 29, and has very much the build of Zepidophora 
lepidocera (appendiculata), except that it is not at all scaly, the 
wings are very much shorter, and so far as can be seen, the 
mouth-parts are not elongated. The anterior tibiae (the only 
ones visible) are not bristly, as they are in Sphenoidoptera. In 
the hairy abdomen it differs from Paracosrmis / but the course 
of the second vein is as in Paracosmus, not as in Metacosmus. 



T. D. A. Cocker ell — Descriptions of Tertiary Insects. 55 

By the characters of the abdomen and venation it is excluded 
from ApKoebantns and the related Epacmus and Eucessia. 

The characters of the venation are as follows : 

(1.) Second vein (R (a+ 3)) arising at the same point as the 
third, and about 390 \x basad of discal cell. Near its end it 
curves upwards, and ends on costa at an angle even more obtuse 
than that presented by Pantarbes (Williston, 1. c, p. 212). 

(2.) Third vein R (4+5) robust, gently curved, with its lower 
branch (R 5 ) forming a gently curved but nowhere angled line; 
upper branch (R 4 ) leaving it at right angles, and after 255 /jl 
bent at a right angle, but emitting a short vein (rudiment of 
the cross-vein) in a straight line with the second section, so that 
the whole branch represents a T resting on the third vein, and 
having one of its sides prolonged and bent upwards to reach 
the costa. The end of the branch is not bent nearly so much 
as the end of the second vein, and its inner angle with the 
costa is not very much greater than a right angle. In general, 
all this is much as in Pantarbes, except as to the rudimentary 
cross-vein directed basad (in Pantarbes the cross-vein is com- 
plete and directed upwards). 

(3.) Radio-medial cross-vein much beyond middle of discal 
cell (about 760 fi from its apex), and 255 \i long. 

(4.) First posterior cell open at apex, the opening about 
120 fj, wide. 

(5.) Discal cell long and rather narrow (its length about 
2f mm ), shaped as in Systrophus (the bounding nervure having 
the same curves), except that it is much longer. 

(6.) Second posterior cell narrow (306 /jl) at base, and 
extremely broad at apex (i. e. on margin) ; third posterior 
broad at base, and narrowed apically ; fourth (morphologically 
fifth) very broadly open, formed as in Pantarbes, but longer ; 
its base is 476 fi from base of discal cell. 

(7.) Anal cell rather widely open. 

Alepidophora pealei sp. no v. 

Length 12 mm ; of wing 6J ; of abdomen about 8 ; width of 
abdomen 3| mm ; head and thorax black ; abdomen dark reddish- 
brown ; wings hyaline, nervures brown, costal region (between 
costa and radius) reddened; as preserved, the abdominal seg- 
ments appear as widely separated dark chitinous bands, with 
broad colorless intervals between, these intervals being about 
half the size of the chitinous rings. Eyes with facets of two 
sizes, the smaller about 22 \x diameter, the larger about 42 /jl ; 
the two kinds seem to be about equally represented, the division 
apparently longitudinal, and not abrupt. Tarsi with scattered 
strong bristles. Third abdominal segment slightly bristly or 
hairy, fourth with a conspicuous patch of dark hairs on each 
side, fifth and sixth moderately hairy, the hairs dark. 



56 T. D. A. Cocker ell — Descriptions of Tertiary Insects. 

Hab. — Florissant in the Miocene shales, 1908. On the same 
slab, close to the fly, are the leaves of Fagus longifolia. While 
I was preparing the above description, Dr. A. C. Peale, the 
discoverer of the Florissant shales, visited my laboratory. It 
is with much pleasure that I dedicate the species to him. 

Holotype in Peabody Museum (Yale). 

Pachysystropus gen. nov. (Bombyliidae). 

Rather large, cylindrical-bodied, with a strong projecting 
proboscis ; antennae witli a sharp apex, constructed essentially 
as in Geron • hind femora stout and hairy. In Williston's 
table (N. Am. Diptera, 3d ed.) it runs straight to Dolichomyia 
(the eyes are not very distinct, but I feel sure not holoptic), 

Fig. 1. 







Pachysystropus rohweri Ckll. 

but the venation agrees better with that of Systropus, differ- 
ing, however, as follows : 

(1.) The lower half of the cross-vein from the bend of R 4 to 
R 2+3 is present, and very distinct, though the upper half is 
wanting. This vein is absent in Systropus (though the abrupt 
bend remains to indicate where it was), but is complete in 
Pantarbes and Exoprosopa. 

(2.) The first posterior cell is closed just before the margin 
of the wing, a condition approximately intermediate between 
that of Systropus (in which it is open) and Pantarbes. 

(3.) The anal cell is closed just before the margin, as in 
Ocondocera (N. Am. Dipt., f. 82). The venation is very well 



T. D. A. Cocker ell — Descriptions of Tertiary Insects. 57 

preserved, and the existence of only three posterior cells is 
certain. The anal cell in the S. African Syst? 7 opus crudelis 
(as figured by Sharp) is like that of P achy systr opus. 

P achy systr opus rohioeri sp. nov. Fig. 1. 

Black or dark brown, the wings hyaline, with dark nervures; 
apical half of area between radius and costa fuliginous ; 
length (excluding proboscis) 13^ mm ; proboscis rather stout, 
a little over 3 mm long ; antennae 2| mm , the last joint very sharp- 
pointed, 663 fi long ; width of thorax 4 mm ; of the parallel-sided 
abdomen about 2| mm , its apex not swollen ; lateral anterior 
corners of first three abdominal segments (especially the sec- 
ond) projecting at an acute angle ; fourth and fifth segments 
with dark hair at sides ; hind femora about 4 mm long, stout, 
the posterior side with much long dark hair ; length of wing 
8f mm ; base to anterior cross- vein 5| mm . 

Hah. — Florissant, in the Miocene shales, Sta. 9. (S. A. 
Pohiver, 1906.) 

Holotype in Peabody Museum (Yale). 

Callimyia (?) hypolitha sp. nov. (Platypezidae). 

A small stout-bodied fly with long wings ; as preserved, the 
head and thorax are dull black ; the abdomen conspicuously 
shining, dark reddish-brown, with the hind margins of the 
segments black ; wings hyaline, nervures pale. 

Length of body 4J mm , of wing 4±- ; length and width of 
abdomen each about 2J mra , but the apex is not visible (appar- 
ently turned downwards) ; thorax robust ; head broad, but not 
quite so wide as thorax. 

Venation, as preserved (the subcosta, radius and branches, 
and anterior cross-vein are visible), exactly as in Callimyia 
(Willst., K Am. Dipt., 3d ed., p. 242, f. 2) except that I 
cannot demonstrate any spinulosity on the radius, but the veins 
are so nearly the color of the rock that this might well be 
invisible. The first posterior cell is widely open, as in Calli- 
myia. The subcosta ends on the costa at about the middle. 

Hob. — Near Rifle, Colorado, in Eocene rocks ; the locality 
and other particulars being the same as already given for 
Tetragon ear a peritula. 

Only two Tertiary Platypericlae have been previously de- 
scribed ; Oppenheimella baltica Meunier, from amber ; and 
Callimyia, torporata, Scudder, from the Green Kiver beds of 
Wyoming. C. hypolitha is a much larger species than C. 
torporata. 

Since writing the above I have found the reverse impression 
of C. hypolitha , showing the anterior branch of the media 



58 T D. A. Cocker ell — Descriptions of Tertiary Insects. 

very distinctly. The media turns downwards near its distal 
end, and the first posterior cell is more widely open than in 
Callimyia. This downward bend is to the apical corner of 
the discal cell, which forms a somewhat acute though large 
angle, and from that point onwards the media is directed 
straight to the margin. (The margin and immediate vicinity 
at this place are not visible.) This arrangement of the media, 
etc. resembles that of Ocydromia (Empididse). Scudder's C. 
torpor ata also has the very widely open first posterior cell. 

Leptis florissantina sp. nov. (Leptidae). 

A small, slender species, beautifully preserved, with expanded 
wings. Length 8 mm ; wing 7 ; expanse 15 mm ; width of head 
2 mm , of the abdomen 1-J-. Antennas not well preserved, but 
the third joint seems to be large, much as in Hilarimorpha ; 
I cannot see the styles. Eyes very widely separated above ; 
head broader than long ; thorax small ; head and thorax black ; 
wings hyaline, without spots, veins dark reddish-brown ; abdo- 
men parallel-sided, a little broadened to the obtuse apex, the 
basal two segments pallid, the others largely dark (perhaps 
originally wholly dark). Venation as in Leptis / compared 
with L. mystacea Macq. (Williston, N. A. Dipt., 3d ed., 
p. 157) the following slight differences are found : 

(1.) Costa not or barely arched near base. 

(2.) Wings narrower. 

(3.) The second vein ends nearer apex of w T ing, beyond level 
of middle of cell in forks of third. 

(4.) Bases of second and third posterior cells about half a 
millimeter back of level of base of cell in forks of third vein. 

(5.) Anterior cross-vein (radio-medial cross-vein) nearer to 
middle of discal cell. The anal cell is barely closed, just on the 
margin of the wing ; there may be an inhnitesimally small 
opening. 

Hob. — Florissant, in the Miocene shales, Station 13 B (S. A. 
Roliwer, 1908). Four species of Leptis have been described 
from amber. 

Tipula heeriana n.n. 

Tipula lineata Heer. Ins. Oen. II, p. 194, t. 15, f. 4 
(1849). — Miocene of Radoboj (not T lineata Scopoli, Ent. 
Cam., p. 320). 

Tipula lineata Heer, Scudder, Proc. Am. Phil. Soc, 1894, 
p. 9. 

Limnophila meunieri n.n. 

Limnophila gracilis (Lw., as Tanysphyra, Scudder, Proc. 
Am. Phil. Soc, 1894, p. 21 (nom. nud.) ; Meunier, Ac. Sc. 
Nat. (9) iv, p. 382, t. 14, f. 9 (1906).— Baltic amber (not 
L. gracilis Wiedemann, Auss. Zw. 1, p. 28). 



Gooch and Beyer — Electrolytic Estimation of Lead. 59 



Art. IV. — The Electrolytic Estimation of Lead and of 
Manganese by the Use of the Filtering Crucible ; by F. A. 
Gooch and F. B. Beyer. 

[Contributions from the Kent Chemical Laboratory of Yale Univ. — cxciii.] 

In a former paper* we have shown that the filtering crucible 
may be put to advantageous use in an electrolytic cell for the 
treatment of deposits not compact and adherent enough to be 
handled rapidly and conveniently by ordinary methods of 
electro-deposition. Four devices were described : I, a closed 
cell in which the perforated platinum crucible is adapted to 
use as an electrode and to nitrations after interruption of the 
electric current ; II, a closed cell in which the perforated 
platinum crucible is used as an electrode and in subsequent 
nitrations without interruption of the current; III, a cell in 
which the perforated platinum crucible is adjusted to a filter- 
ing flask for continuous filtration during electrolysis ; IY, a 
cell in which a perforated porcelain crucible with included 
electrode of platinum is arranged, like the platinum crucible 
of III, for continuous filtration during electrolysis. It was 
shown that by either of the devices described reasonably rapid 
and accurate electrolytic determinations of copper may be 
made without the use of rotating motors or expensive apparatus 
of platinum. In testing these devices, copper sulphate was 
electrolyzed, with an error ranging between + 0*0003 grm. 
and —0*0003 grm. upon approximately 0*1274 grm. of copper 
or 0*5000 grm. of the sulphate. The duration of the elec- 
trolysis was about thirty minutes in the processes involving 
continuous filtration and forty-five minutes in the processes in 
which filtration was begun after electrolysis was completed. 
The metallic copper deposited upon the cathode was partly 
compact, partly spongy. In the work to be described the 
apparatus has been tried in the more difficult determinations 
of manganese and lead as the dioxides formed upon the anode 
in very imperfectly adherent condition. 

The Determination of Lead as the Dioxide. 

In depositing lead dioxidef electroly tically, solutions contain- 
ing nitric acid are employed ; precautions must be taken in 
regard to concentration of acid, strength of current and tem- 
perature ; and the liquid is siphoned off before interruption of 
the current. With the rotating cathode making 600 revolu- 
tions a minute and a sand-blasted platinum dish for the 
anode, Exner obtained in ten to fifteen minutes adherent 

*This Journal, xxv, 249, 1908. 

f Smith, Electro-analysis, p. 101, edition of 1907. 



60 Gooch and Beyer — Electrolytic Estimation of Lead. 



Fig. i. 



deposits with a current N.D. 100 — 10 amp. and 4*5 volts, acting 
upon 125 cmS of solution containing 20 cm3 of concentrated nitric 
acid. 

In some of our preliminary trials of electrolysis in the closed 
cell with subsequent filtration it was found that when the con- 
centration of nitric acid amounted to 10 cm3 in 60 cm3 of liquid, 
with a current of 4 amperes (JST.D. 100 =10 amp.) and 6 volts, 
two sources of error appeared. In the first place, the depo- 
sition of metallic lead upon the cathode was often noticeable ; 
and secondly, it appeared to be impossible to make the precipi- 
tation of lead dioxide complete so long as that substance was 
allowed to float in the liquid. Similar results were obtained 

in experiments in which urea 
was added to the liquid for the 
purpose of obviating the sol- 
vent action of dissolved oxides 
of nitrogen upon lead dioxide. 
In the experiments with this 
form of apparatus the stirring 
of the asbestos felt by gas 
evolved upon the bottom of the 
crucible used as an anode, as 
well as the deposition of oxide 
on the outer surface of the 
crucible, was prevented by tak- 
ing the precaution to moisten 
the asbestos, from the outside, 
with a drop of nitrobenzene 
which, being insoluble in 
water, prevents the contact of 
the aqueous electrolyte with 
the electrode surface under- 
neath the asbestos. An increase 
of nitric acid to the proportion 
of 30 cm3 in 100 cm3 of solution 
served to prevent the deposi- 
tion of lead upon the cathode, 
but to prevent the re-solution of lead dioxide it was found 
to be necessary to use the process of continuous filtration, 
so that the deposit might be compacted upon the felt, and 
after deposition was complete to replace the acid liquid by 
a solution of ammonium nitrate without interruption of the 
current. After washing out the nitric acid with the solution 
of ammonium nitrate the final washing was completed with 
water. The form of apparatus employed, shown in figure 1, 
and the manner of using, were fully described in the former 
article to which reference has been made."'' In Table I are 
* This Journal, xxv, 249, 1908. 




Gooch and Beyer — Electrolytic Estimation of Lead. 61 

given the results of experiments following this procedure, and, 
for comparison, the result of an experiment, (1) in which it 
was found that, though electrolysis was continued by the cir- 
culating process until the filtrate contained no lead, traces of 
lead dioxide went into solution after the amount of electricity 
which passed had been diminished by the gradual dilution with 
water used in washing to replace the electrolyte. Tests for 
lead in filtrates and washings were made by neutralizing with 
ammonium hydroxide and adding ammonium sulphide, or 
acetic acid and potassium chromate. 

Table I. 
Electrolysis with Continuous Filtration. 

HN0 3 Current 

Pb(N0 3 ) 2 Vol. cone. , K ^ Time Pb0 2 Theory Error 

Amp. NDj oo Volt found for Pb0 2 

cm 3 . min. grm. grm. grm. 

A 
With no ammonium nitrate in electrolyte or in wash-water. 
(1) 0-2023 50 15 2 5 4 5 

4 10 5 130 0-1460 0-1436 —00024 

B 

With ammonium nitrate in electrolyte and in wash-water. 



(2) 


0-2022 


50 


15 


2 


5 


4 


40 














4 


10 


5 


100 


0-1459 0-1462 


+ 0-0003 


(3) 


0-2014 


50 


15 


2 


5 


4 


5 














4 


10 


5 


115 


0-1454 0-1458 


+ 0-0004 


W 


0-2001 


50 


15 


2 


5 


4 


5 














4 


10 


5 


115 


0-1444 0-1442 


— 0-0002 


00 


0-2006 


50 


15 


2 


5 


4 


5 














4 


10 


5 


115 


0-1448 0-1446 


— 0-0002 


(6) 


0-2046 


50 


15 


2 


5 


4 


5 














4 


10 


5 


115 


0*1477 0-1472 


— 0-0003 






With 


ammonium nil 


C 
irate in 


wash-water only. 




(8) 


0-2020 


50 


15 


2 


5 


4 


5 














4 


10 


5 


115 


0-1458 0-1460 


+ 0-0003 


(9) 


0-2037 


50 


15 


2 


5 


4 


5 












4 


4 


10 


5 


115 


0-1470 0-1473 


+ 0-0003 



From the results of the experiments described, it appears 
that good analytical results may be obtained with the filtering 
crucible used as an electrolytic cell if nitric acid be present to 
the proportion of 30 cm3 of the concentrated acid in 100 cm3 of 
solution, the liquid kept in continuous filtration until the elec- 
trolysis of the lead salt is complete, the acidic liquid replaced 
by a solution of ammonium nitrate so that the electric current 
passing shall not fall off until the nitric acid has been removed, 



62 Gooch and Beyer — Electrolytic Estimation of Lead. 

the final washings made with water, and the deposit weighed 
after drying at 200°. The time required for the complete 
deposition of 0*15 grm. of lead dioxide under the conditions 
described is about two hours. 

The Determination of Manganese. 

For the electro-deposition of manganese as the dioxide 
various processes have been described.* With stationary 
electrodes solutions containing nitric acid, sulphuric acid, 

Fig. 2. 



* 



J 



acetic acid, formic acid with or without a formate, or ammo- 
nium acetate alone, with chrome alum, or acetone, have been 
employed. For use with the rotating anode, solutions contain- 
ing ammonium acetate with chrome alum or alcohol have 
been advocated. f In all these processes hydrated manganese 
dioxide is deposited upon a large anode which is preferably a 
roughened platinum dish of considerable capacity. 

The experiments to be described have been made to test 
the utility of the electrolytic filtering cell in the determination 
of manganese as the dioxide. The procedure adopted was the 

* Smith's Electro-analysis, edition of 1907, p. 134 et seq. 
f Koester, Zeitsclir. Electro-chem., x, 553, 1904. 



Goocli and Beyer — Electrolytic Estimation of Lead. 63 

simplest. Fifty-centimeter portions of a solution of pure 
manganese sulphate, standardized by evaporation of measured 
portions and gentle ignition of the residue over a radiator,* 
were treated, in each case, with six drops (0'17 cm3 ) of concen- 
trated sulphuric acid and electrolyzed in the filtering cell 
with a current of 2 amperes (JSTD. 100 = 5 amp.) and 20-10 
volts, the voltage decreasing as the solution became heated. 
In one set of experiments the process of continuous filtration 
during electrolysis, for which the adjustment of apparatus is 
shown in figure 1, was employed : in a second set of experi- 
ments the closed cell, shown in figure 2, was used during the 
electrolysis and the adjustment for filtration made subsequently 
as previously described, f The time required for the deposition 
of 0*1860 grm. of the dioxide was one hour and three-quarters 
in the former process : a period varying from two hours and 
ten minutes to two hours and fifty minutes is required in the 
latter process. Tests with hydrogen dioxide and ammonia 
showed that the deposition was complete in the process of 
continuous filtration and practically so in the closed-cell pro- 
cess. The closed-cell process naturally requires less attention 
during the electrolysis, and so it is advantageous to run the 
process for a period, perhaps two hours, with the closed cell 
and then to adjust the apparatus for filtration during further 
electrolytic action in order that floating particles of the dioxide 
may be drawn to the felt and completeness of precipitation 
may be assured. In this way the advantage of the circulating 
process may be obtained with less attention to manipulation 
than is required when the filtration is continuous from the 
start. The deposit was washed with water after interruption 
of the current, first dried at 200° for ten or fifteen minutes 
and weighed, and thereafter ignited to low redness in the 
spreading flame of a large burner 4 Results of experiments 
with the cell arranged for continuous filtration, and of experi- 
ments in which the closed cell was used until the electrolysis 
was nearly over, are given in the accompaning table. 

The results are evidently as good as could be expected of 
any process which involves the weighing of manganese 
dioxide brought to condition by heating. The degree of 
oxidation of the oxide thrown down under the conditions 
which obtain apparently approximates closely to that of the 
ideal oxide represented by the symbol Mn0 2 .H 2 0, formerly 
assigned by Eiidorff§ to the electrolytically formed oxide, 
and differing in that respect from the electrolytically deposited 
oxide which was studied by Groeger.] 

*This Journal, v, 209, 1898. f This Journal, xxv, 249, 1908. 

{This Journal, v, 214, 1898. gZeitschr. angew. Cheni., 1892, 6. 

|| Zeitschr. angew. Chem., 1895, 253. 



6-4 Gooc/i and Beyer — Electrolytic Estimation of Lead. 

Table II. 

Solution H0SO4 Theory Mn0 2 Mn0 2 

of M11SO4 co'nc. Current for weighed weighed Error 

taken , * » Mn0 2 as Mn0 2 as Mn 3 4 

cm 3 cm 3 Amp. ND. 100 Volt grm. grm. grm. 

A 

Electrolysis with continuous filtration. 

(1) 50 0'17 2 5 20—12 0-18G0 0*1862 +0-0002 

0-1858 —0-0002 

(2) 50 0-17 2 5 20—12 0'1860 0'1856 —0-0004 

0-J856 —0-0004 

(3) 50 0-17 2 5 20 — 12 0*1860 0*1843 —0*0017 

0*1872 +0*0012 
B 
Electrolysis in closed cell with subsequent filtration. 

(4) 50 0-17 2 5 20—12 0'1860 0*1860 0*0000 

0-1853 —0-0007 

(5) 50 0-17 2 5 20 — 10 0-1860 0'1856 — 0'0004 

0-1856 —0*0003 

(6) 50 0-17 2 5 18 — 10 0-1860 0'1853 -0-0007 

0-1858 0-0002 

Somewhat greater regularity in results might be expected if 
the manganese dioxide could be converted to manganous sul- 
phate and weighed as such, but experiments made with this 
end in view were unsuccessful. The application of gaseous 
sulphur dioxide proved to be ineffective and sulphuric acid 
attacked the asbestos felt during ignition, causing a permanent 
increment of weight. The substitution of spongy platinum 
for asbestos in the filtering crucible, as suggested and used by 
Monroe,* served to obviate the difficulty due to the action of 
the sulphuric acid upon the filtering medium, but the process 
of removing the necessary excess of sulphuric acid by gentle 
heating over radiator was exceedingly slow and, if pushed, 
liable to errors of mechanical loss. To weigh as either form 
of oxide, in this process, is therefore better than to attempt the 
conversion to manganese sulphate. 

The processes demonstrated for the electrolytic deposition 
of lead dioxide and manganese dioxide with the use of the 
filtering crucible are obviously inferior in point of convenience 
to the more rapid processes which demand the use of large and 
special platinum containers and rotating motors, but in absence 
of such apparatus they may serve a useful purpose. 

* Chem. News, lviii, 101. 



Ashman — Radio- Activity of Thorium. 65 



Art. Y. — The Specific Radio- Activity of Thorium and 
its Products ; by G. C. Ashman. 

The radio-activity of thorium compounds and minerals was 
discovered by Schmidt* and independently by Mme. Curief; 
StruttJ also found that all thorium minerals examined by him 
showed radio-active properties. Debierne§ discovered actinium 
while working up uranium ores containing thorium. The 
similarity of the chemical properties of actinium and thorium 
led to the suggestion by a number of chemists that the activity 
of thorium is not due to thorium itself but to slight traces of 
actinium, since actinium is very difficult to separate from tho- 
rium and the other rare earths. This view is hardly tenable 
in the light of the fact that thorium and actinium give emana- 
tions with totally distinct properties : their rates of decay are 
widely different, and their products are in no way identical. In 
1902 Hofmann and Zerban|| announced that the} 7 had obtained 
an inactive thorium preparation from a Brazilian monazite sand, 
and that thoria prepared from various minerals, although 
active at first, became inactive after a few months. Basker- 
ville and Zerbanl" also claimed they had separated from a 
South American mineral thoria which was perfectly inactive. 
A little later Hahn** and Ramsayff extracted from the min- 
eral thorianite a radio-active preparation several thousand 
times more active than thorium itself and yielding a propor- 
tionately larger amount of thorium emanation. This discovery 
made it appear possible that the activity of thorium might be 
due wholly to radiothorium since the latter possesses an extreme 
tendency to retain its association with that element. Papers bear- 
ing on this question were published simultaneously by Bolt- 
wood,^by Dadourian,§§ and by McCoy and Ross||]|. The con- 
cordant results of these three independent investigations proved 
conclusively that the intensity of the radio-activity associated 
with thorium in any mineral is directly proportional to the tho- 
rium content of the mineral. This work, together with the 
preparation of radiothorium from thorium minerals by Hahn and 

*Ann. d. Phys., lxv, 141, 1898. 

fC. R., cxxvi, 1101, 1898; 

iProc. Roy. Soc., London, lxxvi, 88 and 312, 1905. 

$C. R., cxxix, 593, 1899 ; cxxx, 206, 1900. 

||Ber. d. Chem. Ges., xxxv, 531, 1902. 

IT J. Amer. Chem. Soc, xxvi, 1642 1904. 

**Ber. d. Chem. Ges., xxxviii, 3371, 1905. 

ff J. de chim. phys., iii, 617, 1906. 

itThis Journal, xxi, 409, 1906. 

§§Ibid., xxi, 427, 1906. 

HI Ibid, xxi, 433, 1906. 

Am. Jour. Sci.— Fourth Series, Vol. XXVII, No. 157.— January, 1909. 
5 



66 Ashman — Radio- Activity of Thorium. 

Ramsay, and the demonstration that the extracted radiothorium 
produced thorium-X and the subsequent products left no 
doubt that radiothorium was produced by the disintegration of 
thorium, but the question still remained whether this disintegra- 
tion of thorium itself was accompanied by a-rays ; that is, 
whether thorium itself was radio-active. McCoy and Ross 
found that all of the activity of a thorium mineral, not due to 
uranium, remained in the pure thorium dioxide, separated by 
Neisli's method : the measurements were made at the end of a 
month to allow the accumulation of the maximum amounts of 
thorium-X and subsequent products. Pure thorium dioxide 
made from commercial samples of thorium nitrate had about 
half the normal activity. This was at first thought to be due 
to the removal of part of the radiothorium in the technical 
process of purification. McCoy and Ross attempted to separate 
radiothorium completely from thorium, but found it apparently 
impossible to do so. Hahn found that the activity of com- 
mercial samples varied with the age of the material ; the 
activity was smallest for samples three to nine years old. A 
sample twelve years old had greater activity. To explain these 
results, Hahn* suggested that there is a rayless intermediate 
product, mesothorium, having a period of about seven years, 
between thorium and radiothorium ; that this is separated in 
the process of extraction of thorium from minerals, and that the 
radiothorium, which has a period of two years, decays with time, 
causing the observed fall of activity of commercial samples. 
Some observations by Boltwoodf tended to confirm this view. 
Later quantitative experiments by McCoy and Ross showed 
that the hypothesis was in good agreement with the facts and 
that the period of mesothorium was five and one-half years ; 
this value has since been confirmed by Hahn;); by new quan- 
titative measurements. Hahn also compared the activities of 
old and new thorium by means of the a-rays, the /3-rays, and 
the emanation ; the variation with age in activity of the emana- 
tion is greater than the variation in the a-ray activity. The 
removal of thorium-X caused a greater proportional decrease 
in the activity of new than in old thoria. The activity of 
radiothorium, when freed from thorium-X, was decreased to a 
smaller fraction of its original value than that of any ordinary 
thorium preparation similarly treated. These facts led Hahn 
to conclude that thorium itself has a typical a-ray activity ; 
but he gave no estimate of its intensity. 

Rutherford and Soddy§ had observed that in the case of 
thorium precipitated with ammonia, its final or maximum 

*Phys. Zeitschr., viii, 277, 1907. 
f This Journal, xxiv, 93, 1907. 
JPhys. Zeitschr., ix, 246, 1908. 
§Phil. Mag.,iv, 378, 1902. 



Ashman — Radio- Activity of Thorium. 67 

activity after one month was about four times the initial activ- 
ity. McCoy and Ross* found the ratio of final to initial activ- 
ity to be approximately 2-5 for various samples of thoria 
prepared from a commercial nitrate. The corresponding ratio 
for thoria prepared from minerals was about 3*2. The authors 
at that time considered this to prove conclusively that thorium 
is itself active, and estimated that the specific activity of 
thorium dioxide was between 100 and 130. The work here 
described was undertaken at the suggestion of Professor 
McCoy with the object of determining the a-ray activity of 
thorium itself with accuracy. A brief preliminary mention 
of my results was made in the paper of McCoy and Ross. 

The activity of thorium itself is found from the experi- 
mental data as follows : The activity of new thorium freed 
from the easily separated products thorium-X emanation, 
tliorium A, thorium B, and thorium C, is due to Th -f- Rt. 
Rut in the course of a month the products (Pr) accumulate in 
equilibrium amounts, and the activity is then due to Th + Rt + 
Pr. The activity of old thorium freed from the products 
is due to Th + ccRt, where x is the fraction of the equili- 
brium amount of radiothorium in the sample ; and after an 
interval of one month the activity is due to Th + x (Rt + 
Pr ). These four relations lead to four equations by means of 
which the four unknown quantities, activity of Th, Rt and Pr, 
and the fraction x may be found. McCoy and Ross explained 
that the uncertainty of the data upon which their estimate 
was based was due to the fact that by the large number of 
precipitations with ammonia, which were necessary to remove 
the separable products, Th-X, etc., large quantities of silica 
and other impurities from the glass and reagents were intro- 
duced into the thorium. A further disadvantage of the am- 
monia method arises from the fact that Th-A, with the com- 
paratively long period of 11 hours, is precipitated with the 
thorium. This makes it necessary to carry out 12 to 15 pre- 
cipitations in the course of three or four days in order com- 
pletely to remove the separable products. 

Schlundt and Mooref found that in the precipitation of 
thorium by fumaric acid, according to Metzger's method for 
the analytical separation of thorium from other rare earths, 
Th-A as well as Th-X remains in the solution to a very large 
extent, while Th-B is carried down with the thorium. This 
method would therefore seem to offer an easier way to separ- 
ate the active products of thorium than the precipitation with 
ammonia ; but it had the disadvantage of requiring an expen- 

* J. Am. Chem. Soc, xxix, 1709, 1907. 
f J. Phys. Chem., ix, 682, 1905. 



68 Ashman — Radio- Activity of Thorium. 

sive reagent and the necessity of working in alcoholic solutions. 
Schlundt and Moore* also found that precipitation with nitro- 
benzoic acid removed some form of radio-active matter from 
thorium but they did not identify the products separated. 
The extensive use of meta-nitrobenzoic acid by McCoy and 
Rossf in the analysis and purification of thorium minerals sug- 
gested also the use of this reagent as a substitute for fumaric 
acid. It was found that the first precipitate of thorium with 
meta-nitrobenzoic acid lost its activity at a much greater rate 
than would correspond to the period of thorium- A. The 
minimum activity was reached in four or five hours, then an 
increase due to the growth of thorium-X and its products set 
in. When the precipitate was dissolved and the thorium 
reprecipitated at the end of two hours the activity was dimin- 
ished still further ; and it was found that two additional pre- 
cipitations at intervals of two hours gave finally pure thorium 
dioxide free from thorium-X and all its subsequent products. 
The activity of the oxide prepared by this method began 
immediately to increase. On account of the convenience of 
working in aqueous solutions and the ease and rapidity with 
which filtration takes place, the nitrobenzoic acid method 
offers many advantages over the fumaric acid method, and 
leaves nothing to be desired concerning purity of the product 
and completeness of separation. The precipitation with nitro- 
benzoic acid was carried out in the following manner : The 
solution of thorium nitrate, containing approximately 2 grams 
of thoria, was nearly neutralized with dilute ammonia, methyl 
orange being used as indicator, a slight excess of a saturated 
solution of nitrobenzoic acid was slowly added, and the mix- 
ture kept at a temperature of 80° for a short time. The pre- 
cipitate of thorium nitrobenzoate was filtered from the solution 
and washed well with water. Two hours later it was dissolved 
in dilute nitric acid, diluted to about 300 cc , almost neutralized 
with ammonia and the thorium then reprecipitated by the 
addition of 400 cc more of the nitrobenzoic acid solution. The 
fourth thorium precipitate was rapidly filtered from the solu- 
tion, placed directly in a platinum crucible and ignited first 
over a Bunsen flame and finally, at the highest attainable tem- 
perature of the blast lamp for ten minutes. Four lots of 
thorium oxide were thus prepared ; two from " old " material, 
two from thorite. Oxide "A" was from sample "A" of 
thorium nitrate described by McCoy and Ross.J It had been 
precipitated 100 times with ammonia and was known to be 
poor in radiothorium. It was repurified by the nitrobenzoic 

*Loc. cit. 

f J. Am. Chem. Soc, xxix, 1709. 1907. 

% J. Am. Chem. Soc, xxix, 1712, 1907. 



Ashman — Radio- Activity of Thorium. 69 

acid method and converted into the oxide by ignition. Oxide 
" B " was from sample " B " of thorium nitrate described by 
McCoy and Ross. It had been precipitated 40 times with 
hydrogen peroxide and contained about one third of the 
equilibrium amount of radiothorium. It was twice purified 
by Neish's method, and finally by four precipitations with 
nitrobenzoic acid, made at intervals of two hours, Samples 
a C" and "D'' were oxides made from thorite from Arendal, 
Norway. Nine grams of the finely powdered mineral were 
digested for a number of hours with concentrated hydro- 
chloric acid. The chlorides were dissolved in hot water and 
filtered from the insoluble residue. The thorium was then 
precipitated with a boiling solution of oxalic acid and washed 
free from the excess of the acid. The precipitated oxalates 
were decomposed by boiling them for ten minutes with fifteen 
grams of potassium hydroxide in 30 cc of water. The thorium 
hydroxide was washed free from alkali and dissolved in dilute 
nitric acid. From this solution the thorium oxide was obtained 
free from thorium-X and all subsequent products by four 
precipitations with nitrobenzoic acid in the manner described. 

The thoria thus prepared was made into films, 7 cm in 
diameter, as quickly as possible by Boltwood's method,* paint- 
ing the paste onto a metal disc with a camel's-hair brush. 
Owing to the rapidity with which the activity of the thoria 
increased, the slower sedimentation process, f which yields 
more uniform films, could not be used. However, the error 
due tolack of uniformity was doubtless negligible since one 
set of films, " Dj," made by the sedimentation process gave the 
same maximum specific activity, 948, as did the painted films 
from the same material ; see Table I. Activity measurements 
were made in the manner described by McCoy and Ross,J and 
with the same electroscope. § From five to eight films were 
made from each sample ; the films were made thin, 0*2 to 7 mg. 
per square cm., in order to avoid errors due to the evolved 
emanation in the case of material containing the equilibrium 
amount of Th-X.| 

In order to find the activity of the thorium oxide at the 
time zero, i. e., at the moment when thorium-X and its prod- 
ucts were completely removed from it, the activity of each 
film was measured at intervals of a few hours during the first 
day after its preparation. The increase of activity with time 
was plotted and the activity at time zero found by extending 
the curve back to the ordinate for zero time as illustrated in 

* This Journal, xxv, 269, 1908. 

+ McCoy. J. Amer. Chem. Soc, xxvi, 391, 1905. 

% Log. cit. §This Journal, xxvi, 521, 1908. 

I McCoy and Ross; this Journal, xxi, 438, 1906. 



70 



Ashman — Radio- Activity of Thori 



iu?n. 



the accompanying curve, which represents a film of sample 
" D." The activities are in terms of the U 3 8 standard, which 
had an activity of 35*0 units. This correction was necessary 
as the time intervening between the last precipitation of the 
thorium and the first activity measurement was usually one to 
two hours. 

Fig. 1. 



.55 




10 12 14 16 

Time in Hours. 



18 



20 



24 



Table I gives the results obtained from the four samples of 
oxide. 

Table I. 









Date 


of Measurement 






Date of 


Minimum 




of 


Maximum 


Sample 


preparation* 


activity 


maximum activity 


activity 


"A" 


Aug. 28 


167 




Oct. 1 


400 


«B" 


Oct. 19 


167 




Nov. 21 


380 


"C" 


Aug. 29 


286 




Oct. 4 


921 


"D" 


Oct. 23 


292 




Nov. 23 


948 


"D" 


Oct. 23 







Nov. 23 


948 



All dates refer to the year 1901 



Ashman — Radio- Activity of Thorium. 71 

From these figures the specific activity of thorium alone can 
be calculated. The average value of the specific activity of 
thorium dioxide plus radiothorium and its products given by 
samples " C " and " D " is 939. That of thorium plus radio- 
thorium obtained from the same samples is 289. Therefore 
that of the products alone is 650. From samples "A" and " B " 
the value of the activity of Tha? (Rt + Pr) is found to be 390. 
That of Th + x (Rt) is 107 : therefore xYy is 223 and x = 
0*3446. Since Th + Et = 289 it follows that Kt = 185, and 
therefore the specific activity of thorium dioxide is 104, or that 
of thorium is nearly 119. 

Taxing the period of mesothorium as 5*5 years and that of 
radiothorium as 737 days, McCoy and Ross* have shown that it 
is possible to calculate the activity of a sample of thorium 
dioxide at any time, t, after the separation of all the mesothorium, 
thorium- X and its products. The maximum activity of sample 
"A," Oct. 1, 1907, was 400. On Feb. 27, 1908, 149 days later, 
it was 346. In this interval of time the radiothorium present 
on the former date would decay to e~ 14Q7i * = 0*869, where \ 
is the disintegration constant of radiothorium, and the prod- 
ucts of radiothorium would disappear in the same proportion. 
The specific activity of thorium dioxide itself is 104. If no 
radiothorium had been formed from the new mesothorium 
produced from thorium during the interval of 182 days between 
the date of preparation of the oxide, Aug. 28, 1907, and Feb. 
27, 1908, the activity would have been 0*869 (400-104) + 
104 = 361. The activity of the equilibrium amounts of radio- 
thorium and all its products is 835. Now some radiothorium 
must have been produced from the mesothorium formed from 
thorium during the interval 182 days ; the activity thus gained, 
""according to the formula of McCoy and Ross, would be 

836 | 1-1"" ^ 2 (e~ A ^ - e~ x A "1 - e~ M 1=4. 

The calculated specific activity on Feb. 27, 1908, is therefore 
361 + 4 = 365. The observed activity was 346. Table II 
gives the activities observed and calculated of samples "A," 
" B " and " D " on various dates since their preparation. 

Table II. 
Date of 
Sam- prepar- Date and max. 
pie ation activity Date and activity Date and activity 

1907 1907 1908 Obs. Gal. 1908 Obs. Gal. 



"A" Aug. 28 Oct. 1 400 Feb. 27 346 3G5 Aug. 26 308 336 

"B" Oct. 19 Nov. 2 1 380 Feb. 27 346 359 Aug. 29 308 331 

"D" Oct. 23 Nov. 23 948 Apr. 841 854 Aug. 28 767 765 

* Loc. cit. 



72 Ashman — Radio- Activity of Thorium. 

It is seen that in the case of sample " D" the observed and 
calculated activities agree quite closely. The discrepancies are 
probably due mainly to errors of experiment. In samples "A" 
and " B " the calculated activity is somewhat greater than the 
observed activity, for which there is no adequate explanation. 
The quantitative changes with time in the activity of thorium 
and its products would be effected to a slight extent by the 
presence of actinium or ionium* in the thorium extracted from 
uranium-bearing minerals ; but, as the percentage of uranium 
in the mineral used was not determined, the magnitude of 
this effect can not be calculated. 

The more important results of this investigation may be 
given as follows : An expedient method for separating thorium 
and radiothorium from the other disintegration products of 
the element has been described. Additional evidence has been 
obtained which supports the view that thorium on disintegrating 
emits an a-radiation. The specific activity of thorium is about 
119 ; this is 11 per cent of the activity of the same thorium 
when equilibrium amounts of its disintegration products are 
present. Of the remaining activity 20 per cent is due to 
radiothorium and 69 per cent to Th-X, and subsequent prod- 
ucts. The activity ascribed to thorium itself is too much in 
excess of the activity of the ionium associated with thorium in 
minerals to warrant the conclusion that the observed activity 
of thorium is due wholly or largely to the presence of ionium. 
The earlier conclusions of Hofmann, and of Baskerville and 
Zerban, that thorium itself is inactive, were without doubt due 
to the fact that the methods used were not sufficiently refined 
to detect the low activity of the element free from its short-lived 
active products, and other radio-active impurities. 

* Boltwood, this Jour., xxv, 269, 1908. 

Kent Chemical Laboratory, 
University of Chicago, 
October, 1908. 



C. Barus — Coronas with Mercury Light. 73 



Art. VI, — Coronas with Mercury Ligjit ; by C. Barus. 

1. Preliminary Survey. — The inferences of the preceding 
papers* gave promise that on judiciously using monochromatic 
light as the source of illumination, the optical nature of the 
coronas might be fully brought out. Such light must be strictly 
homogenous and at the same time very intense. Hence the 
usual methods of obtaining it are unsatisfactory. The strong 
green line of a mercury lamp, however, fulfills the require- 
ments admirably, and this was therefore used. The results 
show that the green disc and the first green ring alternately 
vanish as the result of the interference phenomenon superim- 
posed on the diffraction phenomenon. If therefore the nucle- 
ation of a highly charged medium is systematically reduced, 
a series of angular diameters may be obtained both for the 
green disc and the inner or outer edge of the first green ring. 
From the loci of these values the position of the first diffrac- 
tion minimum for green light may be inferred, and the size 
of droplets computed from the usual equation for small opaque 
particles. 

If the reduction of the nucleation is accomplished by succes- 
sive partial exhaustions, all of them identical, while filtered air 
is allowed to enter the receiver systematically between the 
exhaustions, the nucleations of any two consecutive exhaustions 
should show a constant ratio. Allowances must, however, be 
made for the subsidence during the later fogs and for time 
losses, if any. This is the method used hitherto in my work 
and the results seem to have been trustworthy. 

In the case of mercury light, however, it is now possible to 
compare the latter with the former (diffraction) method of 
obtaining the diameters of particles, with a view to throwing 
definite light on the optical phenomenon. Subsidence methods 
are out of the question for large coronas, as these are invariably 
fleeting in character and pass at once into smaller coarse 
coronas. 

The results of the two methods may be regarded as coinci- 
dent as long as not more than 300,000 nuclei per cubic centi- 
meter, or diameters of particles not smaller than -0003 cm are in 
question. For larger numbers and smaller diameters the 
divergence rapidly increases. Indeed, for particles larger than 
the size given, the optical loss per exhaustion exceeds the 
exhaustion ratio, a result which is satisfactorily explained by 
the cotemporaneous subsidence of these relatively large par- 
ticles. For particles smaller than the limit in question, the loss 

* This Jour., xxv, p. 224, 1908 ; xxvi, p. 87, 1908; xxvi, p. 324, 1908. 



74 C. Bar us — Coronas with Mercury Light. 

computed by the optic method is larger than the exhaustion 
loss, as if fresh nuclei were produced, or rather made available 
at each exhaustion. It is this result which .the present paper 
purposes to bring out in detail and to consider in its bearings 
on the optical phenomenon. 

2. Apparatus. — The fog chamber was of the usual pattern, 
cylindrical in form> with its axis horizontal. The clear walls 
being of blown glass showed some refracting disturbances, not 
however of serious character. The fog chamber was connected 
with a large vacuum chamber by a short, wide passage way, 
though width is of little consequence here. The cylinder was 
lined with wet cloth, closely adhering, except at the narrow 
horizontal windows for observation. 

For exhaustion this stopcock was suddenly opened at the 
beginning of the first second, closed after five seconds and the 
corona quickly measured. Filtered air was then at once intro- 
duced and the next exhaustion made at the beginning of the 
sixtieth second. This rhythm is essential. The isothermal 
value of a drop of pressure [Sp^] was carefully predetermined. 
It fixes the ratio, y, of the geometric progression of nucleation, 
since 

y = (^-tt-^J) / (p-tt) 

where p is the barometric pressure and it the vapor pressure 
at the given temperature. If the cock were left open for a 
longer time than five seconds, [SjpJ would increase to the limit 

The goniometer was of the usual type, the eye being at the 
center or apex, while two needles on radii 30 cm long registered 
the angular diameter of the coronal discs or annuli. Formerly 
the whole instrument was placed on the near side of a fog 
chamber, the eye being about 30 cm from the nearest wall. 
It conduces to much greater sharpness of vision, however, 
and admits of a measurement of larger coronas, caet. par.. 
if the eye is placed all but in contact with the nearer wall and 
the needles (or in this case preferably the inner edges of round 
rods) beyond the further wall. In such a case the refraction 
errors are also diminished. In addition to these advantages 1 
may mention the decidedly increased (about 25 per cent) value 
of the aperture obtained. These excessive apertures show, 
however, that the ordinary diffraction equation for coronas 
is not fully applicable ; for aperture varies with the position 
of the eye along the line of sight. It is often surprising how 
large a corona can be measured by the second method in a 
small fog chamber, scarcely six inches long. The distance 
between lamp and chamber is kept about D = 250 cm . 

3. Equations. — The equations needed in the present work 
are derived in my last Report,* and need merely be summarized 

* Carnegie Publications, No. 96, 1908, Chapters I, III (equat. 1 to 12). 



C. Barus — Coronas with Mercury Light. 75 

here. If y is the exhaustion ratio, the nucleation n z of the 
2th exhaustion in terms of the original nucleation n will be 

n, = n.y(i-,s/0(i-s/0 — -(i-*/**-) 

where S is a subsidence constant and s the chord of the 
angular diameter of the coronal disc on a radius of 30 cm . 

To find jS, two consecutive values of s suffice, or 

S=s z *(l-s z 3 +1 /ys z 3 ) 

approximately. The diameter of particles, d, is in terms of n, 

d 3 = 6m/irn 

If X = 54*6 X 10~ 6 cm. is the wave length of green mercury 
light and 6 the angular radius of the first green minimum of 
the coronas 

sin 0=-61 X/ (d r /2). Since sin 6 = s / 2JR, 

R being the radius of the goniometer of which s is the chord, 
d' the diameter of fog particle (the primes referring to optic 
measurements) 

d'= -004 /s 
for mercury light. Hence, optically, the nucleation 

1(T 6 n' = 29-8 ms\ 
where m grams of water are precipitated per cubic centimeter 
on exhaustion, and found in the exhausted fog chamber. 

4. Data with green mercury light, 10 6 \ = 54'6 cm . — Omitting 
the correlative data with white light and removing all tables 
for lack of space, the charts figure 1 contain results obtained 
with a mercury arc lamp as the source of light. Different 
annuli are measured and the chords, s, on a radius of 30 cm are 
distinguished by accents, as follows : 

s f is the chord of the green disc, 

s" is the chord of the inner edge of the first green ring, 

s" r is the chord of the outer edge of the first green ring. 

The edges are fairly sharp. Hence 
s- (s' + s")/2 

may be taken as the chord of the first green minimum. Optic- 
ally the diameter of the particle is then d' — '004/5. 

The water precipitated per cubic centimeter, m, differs with 
the drop of pressure, $p 9 /p, and the temperature. In the suc- 
cessive parts of the chart the data are 

Parts I, II, 10 8 ?n=4-l grams, »'=122 s 3 (omitted) 

Parts III, IV, =4*12 grams, ^' = 123 s 3 

Parts V-VIII, =4-02 grams, n' = 120 s 3 



76 



C. Bants — Coronas with Mercury Light. 



where n' \& the optical value of the number of nuclei per cubic 
centimeter. 

In parts I-IV the angular diameter, s', of the green discs 




only were measured. The diameter s of the corresponding 
first minims may, however, be obtained by using the method of 
reduction found in parts V and VI where s='44+ '85 s'. 
In parts V and YI the data for s' and s'\ the angular diam- 



C. Barns — Coronas with Mercury Light. 77 

eter of the inner edge of the first ring, are both observed, while 
in parts VII and VIII data for s' and s'", the outer diameter 
of the first ring, appear. 

In parts I and II (omitted) the goniometer was in front of 
the fog chamber, and in series I a two-minute interval between 
exhaustions was (exceptionally) introduced. The result is not 
good ; for a sudden break of the curve appeared after the 
seventh exhaustion, probably due to the time losses in the extra 
minutes. The reason, however, is by no means obvious. In 
series II, for one minute intervals there was no break and the 
locus passing through the points for green discs (the others, not 
marked g, are here and elsewhere to be disregarded) is per- 
sistently straight throughout. The curves show 

For series I, ds / dz = 1 '00, and 
For series II, ds / dz = '80, 

suggesting a time loss in the first case. 

Compared with the preceding table, the values of ds/dz 
should be in the ratio of red and green minims, or 

ds r J dz : ds g /dz = *95 / -80 correspond to A r /\ = 63 -0 / 54'6 ; 

the last ratio, 1*15, is somewhat short of the former, 1*19. 

In series III and IV the pins of the goniometer are behind 
the fog chamber, the eye being at the front wall. In series III 
the relation, 

/ = 2-30 + 1-15 (19—2) 

is remarkably well sustained throughout ; and in series IY 

s' = 2-50 + 1-13 (18 — z) 

gives a good account of the green coronas, if the dull cases are 
ignored. 

The most interesting results are given in parts Y and YI of 
figure 1, and in the subsequent paragraphs the computations 
have been fully carried out. In these cases the chords on a 
radius of 30 cm of the edge of the green disc, s', and the inner 
edge of the first green ring, s'\ were successively observed. 
Figure 1 contains both pairs of curves and their linear character 
is again astonishing. We may write 

Part V, s 7 =2-0-f 1-10 (19 — 2), Part VI, s'=2-0 + l*13 (18 — z), 

s"=3-4 + l-21 (19—2), s" = 3-3 + l-30 (18-3), 

s= -59 + 1 '06s' = — -56 + -96s", s= '50 + 1 -07s' — — -44 + -93s", 

where the minimum is located midway -between s' and s", both 
of which are fairly sharp. 

From both series the mean value 

s = -55 + 1 -06 s'= — -50 + -93 s" 



78 C. Barns — Coronas with Mercury Light. 

may be derived, for the general reductions in this and other 
cases where s f is observed. The individual data are omitted. 

With the given value for s the optical data for the diameters 
of the particles, d'='0Q4:/s and for the nucleation, ?z/ = 120 s% 
were computed. 

The subsidence constant 

S = s z > (l-s^Jys'J 

was obtained from the observations between 2 = 13 and z = 19, 
and a mean datum, #=12, accepted as most satisfactory. It is 
then possible to compute n 09 the original nucleation, as 

n = nj \y* (1-S/O - — (1^-S/*. 1 -,) 5 
from each of the groups of values specified. The mean results 
are 

Part V, rt =4 540 000 

Part VI, =3 430 000. 

Knowing n , the series of data for the nucleation n X 10~ 3 
follow ; from these the diameter 10W=199 ?i-b of fog particles 
and the apertures s = *004/^ are finally computed. In other 
words n and d are the data obtained in view of the occurrence 
of geometric series of nucleations with allowance for subsidence. 

All these results have been constructed graphically in figures 
2 and 3, for parts V and VI of the above charts respectively. 
The observed and computed apertures s are first given ; then 
the observed are optically computed d' and the computed d from 
sequences appear, and finally the observed or optic n' and the 
computed nucleation n from sequences. The corresponding 
data coincide very closely after 9 to 11 exhaustions, i. e. for 
moderately small coronas : but in the case of large coronas 
and nucleations, there is marked systematic divergence. A 
discussion of these results will be given presently. 

Finally, parts VII and VIII of fig. 1 contain measurements 
of the chord of the angular diameter (radius of goniometer of 
30 cm ) of the outer edge of the green disc s' and the outer 
edge of the first ring s' r/ . The minimum may be deduced from 
s\ since s = -54+1 -06s' has been accepted. From s the optic 
value of diameter of fog particle d' and the nucleation n' 
have been computed. The apertures are very fully given in 
figure 1, in relation to the number z of the exhaustion. The 
curves are again linear in character, but in case of the outer 
edge s'" 9 the insufficient size of my fog chambers and the 
greater vagueness of definition interferes with close measure- 
ment. The results are 

Part VIIIs' =1-00 4- P07 (I8-2), part VII •*'= -50 + 1-13 (17— 2), 
s'"=:3-00 + l-93 (I8 — 2), s"'=l-50 + 2-13 (17-2), 

s =1-61 + 1-14(18-2), s .-=1-08 + 1-20 (17 — 2). 



C. Barus— Coronas with Mercury Light. 79 

As a rule s'" is less than twice s' '. The data for s' and s are 
obtained alternately as the figure shows, the green color passing 
from the disc of the first ring and returning again, in suc- 
cession. 




Inferences. 

5. Interference and diffraction. — The tables and chart show 
in the first place, that the disc and first ring of the coronas are 
alternately of a vivid green, other colors being dull because the 
remaining lines of the mercury spectrum are faint. At inter- 



80 C. Barus — Coronas with Mercury Light. 

vals neither disc nor ring are quite green. Hence there is a 
periodic term impressed on the diffractions, which may be iden- 
tified as an interference similar to the case of the lamellar 
grating referred to in another paper.* 

When monochromatic light is used it is necessary, therefore, 
to observe both the edge of the disc and the inner edge of the 
first ring ; for neither appear vividly at the same time. The 
chord s on a radius of 30 cm for the minimum in terms of 
the corresponding chords s' and s'\ of the first and second 
edges specified, may then be written. 

s = -55 + 1*06 s', 
s = — -50 + '93 s" 

Probably the ratio s/s' and s/s" should be constant and the 
absolute term in these equations is an error of observation ; but 
as it is small, little depends upon it, millimeters only being 
significant. 

If we summarize all the observations for ds/dt, the agree- 
ment as a whole is in keeping with the nature of the observa- 
tions and reasonably satisfactory. Thus, in 

Series I, II, d'sjdz = *90, } s = '44 '-[- *85 s' (goniometer in front), 

III, 1-15,1 

IV, 1-13, I 

Y, 110, ! s= -54 + 1*06 s' (goniometer behind), 

VI, 1*13, [ mean ds r /dz = 1-12. 

VII, ' 1-13, | 

VIII, 107, J 

The feature of these data is the occurrence of linear loci for 
s and nearly throughout the extent of the curves. It is as 
difficult to even conjecture a reason for this, as it is easy to find 
reasons against it. The presumptive equation for s is 

s = -004 (7r/6m)* n*, oc 3y^ 

and for n' z if II denotes the product, 

nz = n Q y z Ii. 
Hence, 

s == -004 (7m /6m)* (y z II)^ 
if we disregard the subsidence correction for large coronas 

ds I dz oc y z/3 

in which there is no suggestion of a sustained constancy of the 
coefficient ds/dz such as the experiments show. 

To come to some conclusion as to the cause of the discrepancy 
between the optic value of the nucleation n' and the presum- 
able value n (geometric progression), we may compare the 
* This Journal, xxv, p. 224, 1908. 



C. Barus — Coronas with Mercury Light. 81 

successive value n' zJr Jn f z — s 3 z+1 /s\ in their relation to y — -771 
the exhaustion applied. • The table shows that for very large 
coronas n / z+1 /n / z >y, whereas for very small coronas n f z+ Jn' z <iy. 
For the intermediate coronas (g 5 ), i. e., from the seventh to the 
tenth exhaustion among twenty, the ratio is nearly equal to y. 
Ratios smaller than y may be reasonably interpreted as due to 
subsidence and the constant Sis actually of the order of values 
to be computed from the viscosity of the medium and the size 
of the vessels and fog particles. Within this range (coronas 
smaller than g 9 ) the optic and the presumptive nucleation may 
in fact be brought into agreement. 

In case of the large coronas, however, subsidence is virtually 
absent and the occurrence of n r z+1 /n z > y calls for some appar- 
ent production of nuclei at each exhaustion, which is altogether 
improbable. Therefore, fig. 2 and 3, compare n' and n, whence 
n-n' shows the number of nuclei not registered by condensation. 

For, no matter whether condensation on a given group of 
nuclei occurs or not, no matter how many nuclei have failed of 
catching a charge of water, the identical removal of nuclei by 
partial exhaustion must take place. Such removal is indepen- 
dent of condensation, and would occur in a dry atmosphere 
under similar treatment. Consequently y cannot be too large. 
It may be too small not only from subsidence, but from time 
losses (decay), or as the result of the purification of air due to 
turbulent motion across a solid or liquid surface. Consequently 
n may be regarded as an inferior limit of the nucleation with 
a probably close approximation to the true value. A comparison 
of n and n-n' would, in such a case, show the percentage of 
nuclei of irregular size which have failed of capture, .the 
number being n-n'. 

At the same time it must always be recalled, that no ade- 
quate theory of coronas exists and that therefore the meaning 
of n' is obscure. We must in any case place a part if not all 
the discrepancy between n and n' within the province of such 
a theory as is evidenced by the dependence of aperture on the 
position of the eye. The need is particularly manifest for the 
large coronas, in which there is accentuated superposition of 
interference and diffraction. Small coronas may be tested by 
coincident results obtained from subsidence and the agreement 
is then well within the errors of observation. 

Brown University, 

Providence, R. I. 



Am. Jour. Sci. — Fourth Series, Vol. XXVII, No. 157. — January, 1909. 



82 Scientific Intelligence. 



SCIENTIFIC INTELLIGENCE. 



I. Chemistky and Physics. 

1. An Attempt to Produce a Compound of Argon. — Assum- 
ing that any compound which argon might form would be 
endothermic, and would hence require a large amount of energy 
and a high temperature for its production, and assuming also that 
such a compound would need to be quickly cooled in order to 
prevent its decomposition, Fischer and Iliovici have passed elec- 
tric sparks, as well as the electric arc, between poles of cadmium 
in liquid argon. Cadmium was selected for these preliminary 
experiments on account of the fact that Kohlschtitter and M tiller 
had found that the pulverization of cadmium by the electric 
discharge was abnormally great in argon gas, as compared with 
other gases. The argon was prepared from atmospheric air by 
the absorption of the other gases by means of calcium carbide. 
An elaborate apparatus was devised for carrying out the experi- 
ments, in which it was found necessary to keep the argon at a 
pressure near that of the atmosphere, for liquid argon boils at 
— 189*6° C. under a pressure of 760 mm , while it solidifies only 
2*70 below this point, and below a pressure of 500 mm liquid 
argon does not exist. As a result of the passage of electric 
sparks and the arc through this liquid there was produced a new 
compound, cadmium nitride, due to traces of nitrogen in the 
argon used. This nitride, which was mixed with pulverulent 
metallic cadmium in the product produced by the arc, gave off 
nitrogen when heated in a vacuum, and this nitrogen was found 
to contain a considerable amount of argon. The authors con- 
sider this argon as due to adsorption, but propose to continue 
their experiments under different conditions. — Berichte, xli, 3802. 

h. l. w. 

2. Hhiplosive Crystallization. — Haying occasion to evaporate 
a solution containing a sulphate and a thiosulphate enclosed in a 
bell-jar under a diminished pressure of about 20 mm , Weston 
found that an apparent explosion violent enough to break the bell- 
jar had occurred during the night, when the liquid had become 
very concentrated. Soon afterwards he attempted to crystallize 
two solutions of a sulphite, which were probably supersaturated, 
under the same conditions. In this case he saw a part of the 
contents of one of the dishes violently ejected so that the dish 
was broken, the other dish was upset, and the whole apparatus 
was so shaken that the glass plate upon which the bell-jar rested 
was broken. The author is of the opinion that crystallization 
was suddenly induced on the surface of the basin with a con- 
sequent sudden increase in the vapor-pressure of the surrounding 
liquid, which under the very low pressure existing in the bell-jar 



Chemistry and Physics. 83 

caused the liquid to boil violently. It appears to the reviewer 
that this explanation overlooks the chief cause of the sudden 
boiling, namely, the great amount of heat that is set free by the 
sudden crystallization of a large amount of salt from a supersatu- 
rated solution. This sudden warming under low pressure and the 
presence of much solid salt in the liquid explain satisfactorily 
these interesting accidents. — Chem. JVews, xcviii, 27. h. l. w. 

3. Constituents of Ytterbium. — Aiter von Welsbach, whose 
work on the rare earths is well known, particularly his splitting 
up of old didymium into neodymium and praseodymium, has suc- 
ceeded in obtaining two earths from ytterbium. This was done 
by long continued fractionation, particularly of the ammonium 
double oxalates. He has named the new elements, Aldebaranium 
and Cassiopeium, Ad and Cp. Their distinguishing feature is 
in their spark spectra. Their reactions are those of ytterbium, 
and they cannot be distinguished from one another by any 
chemical reactions. They form only one oxide, the sesquioxide, 
which is stable at a red heat. All the salts are colorless if the 
acid is not colored. The atomic weights found were 172*90 for 
Ad and 74-23 for Cp. The cassiopeium was not obtained abso- 
lutely free from the other earth, but it is the author's intention 
to repeat the separation with a larger amount of material in order 
to obtain purer cassiopeium, which he designates as the last in the 
series of rare earths. It will take six or eight years to complete 
this tedious task.— 3Ionatshefle, xxix, 2. H. l. w. 

4. A New Form of u Tin Infection." — It has been known for 
a long time that metallic tin at low temperatures is subject to a 
physical change which renders it grey and very friable. For 
instance, it is said that the tin buttons on the uniforms of 
Napoleon's soldiers in the Russian campaign fell to pieces. This 
behavior was formerly attributed to the simple action of low tem- 
perature, but in recent times it has been shown that it is due to a 
recrystallization which can be communicated to any piece of tin 
at a low temperature, by contact with an affected piece of tin, or, 
as it may be termed, by inoculation. It has been recently observed 
by von HAssLiNGERthat a piece of tinware soldered with tin which 
had been stored for two years in a place which was alwaj^s heated 
in winter showed a dull surface, a crystalline structure under the 
microscope, and a friability corresponding to the change that 
usually takes place only at low temperatures. It was found 
further that this material was capable of inoculating unchanged 
tin, either in the form of ordinary tin plate, tin foil or cast tin, at 
such temperatures as 7°, 19° and 37° C. The growth of the 
infected spots of nearly circular form was at the ra>e of from 3 
to 5 mm per day, but it was noticeable that this growth became 
slow as it extended farther from the inoculated center. — Monat- 
shefte, xxix, 787. h. l. w. 

5. The Heat Evolved by Radium. — A new determination of 
this constant has been made by von Schweidler and Hess, with 
the result that 118*0 gram-calories per hour were found. Previ- 



84 Scientific Intelligence. 

ous observers have obtained results varying from 100 to 134 in 
the same terms. The material used consisted of l*0523g. of radi- 
um-barium bromide, corresponding to -795 lg. of radium. The 
method used was that previously employed by Angstrom for the 
same purpose. It consisted in placing the radium in a calorimeter, 
and keeping a second, exactly similar, calorimeter at the same 
temperature by an electrically heated wire of known resistance. 
The elevation of temperature thus produced in the calorimeters 
amounted to 5 -5° C. At first only 40 per cent of the expected 
heating was produced, probably on account of the escape of some 
emanation before the beginning of the experiment, but this 
increased from day to day. The radium employed was in equi- 
librium with its decomposition products of short period, but its 
contents of the products from radium D to radium F was uncer- 
tain. — Monatshefte, xxix, 853. h. l. w. 

6. Positive Rays. — W. Wien arranged an apparatus which 
enabled him to transmit the canal rays some distance through a 
capillary tube which was exhausted by means of charcoal and 
liquid air to a low vacuum. He concludes from his observations 
that the particles in the canal rays which are least influenced by 
a magnetic field are those which carry the light emission. More- 
over, the condition of equilibrium of particles of equal weight — 
which may be destroyed by a magnetic field — reasserts itself dur- 
ing the length of path of the rays. The length of this path 
increases with higher potentials. — Physik. Zeitschrift, No. 22, 
Nov. 1, 1908, pp. 765-767. J. t. 

7. Spectral Intensity of Canal Pays. — The observers of the 
Doppler Effect in canal rays in hydrogen agree that the displace- 
ment observed is a band which is separated from the hydrogen 
line by a space showing no band. J. Stark and W. Stenberg 
select the method of viewing the hydrogen line at right angles 
to the direction of the canal rays, in order to determine its 
changes in intensity. No Doppler effect of course is seen, but 
the photometric measures indicate the changes in velocity of the 
positive particles. They state that the intensity of the canal 
rays depends upon their velocity. They also find a minimum 
intensity in the Doppler effect. — Ann. der Physih., vol. xxvi, 
1908, pp. 918-926. . J. T. 

8. Canal Rays. — J. Stark discusses the bearing of modifica- 
tion of the electromagnetic theory of light upon the phenomena 
of canal rays, especially Planck's theory of electrical resonators, 
and Einstein's discussion of lichtquantenhypothese, and believes 
that this latter hypothesis must stand or fall with further observa- 
tions of the canal r&ys.—Physik. Zeitschrift, No. 22, Nov. 1, 
1908, pp. 767-773. j. t. 

9. Potential Measurements in the dark Cathode Space. — W. 
Westphal has confirmed Prof. J. J. Thomson's theory that the 
cathode rays are due to impacts of positive particles impinging 
through the cathode space upon the cathode, and shows, also, 
that measurements of this potential by means of sounding wires 



Chemistry and Physics. 85 

does not introduce an appreciable error. He finds that Poisson's 
equation is applicable to electric discharges through a gas. — Ann. 
der Physik, No. 13, 1908, pp. 571-588. " J. t. 

10. The Elements of Physics ; Vol. II, Electricity and Magnet- 
ism ; by E. L. Nichols and W. S. Franklin. New edition, pp. 
vii + 303, with 196 figures. New York, 1907 (The Macmillan 
Co.). — This edition of the second volume of Nichols' and Franklin's 
Elements of Physics has been entirely rewritten. No statement 
is made here, however, concerning a new edition of the other 
volumes, nor have we seen any announcement elsewhere. 

The most salient feature of the new edition is the discarding 
of the traditional treatment of electrostatics, beginning with elec- 
trostatic attraction and the definition of the electrostatic unit of 
charge. The authors say: "It seems better to approach this sub- 
ject from the standpoint of the ballistic galvanometer, inasmuch 
as, when so developed, the theory of electrostatics is a logical con- 
tinuation of the foregoing theory of the electric current." "Most 
students begin electrical theory at both ends and never reach the 
middle." 

The columns of mechanical and electrical analogies at the end 
of Chapter VII and at the beginning of Chapter XVI are inter- 
esting and complete. At the close of the book a list of 145 well- 
selected problems is given. The volume as a whole is attrac- 
tively executed and reflects credit both on the authors and on the 
publishers. h. s. it. 

11.-4 Text-Book of Physics ; edited by A. W. Duff. Pp. xi + 
680, with 511 figures and 225 problems. Philadelphia, 1908 (P. 
Blakiston's Son & Co.). — This volume is a new departure in the 
writing of college text-books, in that seven experienced teachers 
have contributed to its production. The various divisions of the 
subject and the authors responsible for them are as follows : 
Mechanics, pp. 1-177, written by A. W. Duff; Heat, pp. 179-281, 
by K. E. Guthe; Wave Motion and Sound, pp. 283-328, by W. 
Hallock; Light, pp. 329-474, by E. P. Lewis; Electricity and 
Magnetism, pp. 475-630, by A. W. Goodspeed ; Conduction of 
Electricitv through Gases and Radio-Activity, pp. 631-666, by 
R. K. McClung. 

An obvious danger attendant upon each contributor writing 
about his special field is that of his treating the subject at such 
length as to weaken in the student's mind the contrast between 
the fundamental facts and those of less significance. In other 
words, too much detail tends towards a dead level of the special 
branch under discussion. This danger does not seem to have 
been successfully avoided, in spite of the fact that the sections 
written by each author were freely criticized by his six collabor- 
ators. This objection is especially pertinent to the chapters on 
Light. In the preface the suggestions are made: "Some may 
find the material included in the book too extensive for a single 
course. If so, a briefer course can be arranged by omitting all 
of the parts in small print together with as much of those 



86 Scientific Intelligence. 

in large print as may seem desirable." This is unquestionably 
true from the theoretical standpoint, but we have found by 
experience that, in general, much culling out of paragraphs is not 
conducive to the best results on the part of the student. 

The processes of the differential and integral calculus are used 
only in the small-print paragraphs and the notation of infinitesi- 
mals occurs very infrequently in the large-type articles. Never- 
theless, several useful formulae, which in their less rigorous, 
algebraic clothing are very useful in numerical examples, occur 
only in the small-print paragraphs in conjunction with the more 
rigorous, calculus expressions for the same laws. 

The problems are not numerous and are grouped at the ends 
only of the chief divisions of the entire subject to w^hich they 
appertain. The percentage of rather poorly drawn, shaded 
figures is appreciably greater in this volume than is usually the 
case with recent books of college grade. • h. s. u. 



II. Geology. 

1. Publications of the United States Geological Survey, 
George Otis Smith, Director. — Recent publications of the IT. !S. 
Geological Survey are noted in the following list (continued from 
p. 402 of vol. xxvi) : 

Topographic Atlas. — Thirty-five sheets. 

Folio No. 161. Franklin Furnace Folio, New Jersey. Descrip- 
tion of the Franklin Furnace Quadrangle ; by A. C. Spencer, H. 
B. KfJMMEL, J. E. Wolff, R. D. Salisbury, and Charles 
Palache. Pp. 27, with 6 maps, columnar sections, 15 figures. 

Bulletins. — No. 347. The Ketchikan and Wrangell Mining 
Districts, Alaska; by Fred Eugene Wright and Charles 
Wtll Wright. Pp. 210 ; 12 plates, 23 figures, 3 maps in pocket. 

No. 349. Economic Geology of the Kenova Quadrangle, Ken- 
tucky, Ohio, and West Virginia ; by William Clifton Phalen. 
Pp. 158, 6 plates, 21 figures. 

No. 351. The Clays of Arkansas ; by John C. Branner. 
Pp. 247 ; 1 plate, 20 figures, 1 map in pocket. 

No. 352. Geologic Reconnaissance of a Part of Western 
Arizona ; by Willis T. Lee. With notes on the Igneous Rocks 
of Western Arizona; by Albert Johannsen. Pp. 96 ; 11 plates, 
16 figures. 

No. 355. The Magnesite Deposits of California ; by Frank 
L. Hess. Pp. 67, 12 plates, 4 figures. 

No. 357. Preliminary Report of the Coalinga Oil District, 
Fresno and Kings Counties, California ; by Ralph Arnold and 
Robert Anderson. Pp. 142, 2 plates, 1 figure. 

No. 362. Mine Sampling and Chemical Analyses of Coals 
tested by the United States Fuel-testing Plant, Norfolk, Va., in 
1907 ; by John Shober Burrows. Pp. 23. 



Geology. 87 

No. 369. The Prevention of Mine Explosions, Report and 
Recommendations; by Victor Watte yste, Carl Meissner and 
Arthur Desborough; with letter of transmittal by J. R. Gar- 
field. Pp. 11. 

Also 341-A. Advance chapter on the Coal Fields of North 
Dakota and Montana from Bulletin 341, Contributions to Econ- 
omic Geology, 1907, Part II. 

Further, Mineral Products of the United States, 1896-1907. 
Tabulated on large sheet ; also numerous advance chapters from 
Mineral Resources of the United States, 1907 ; and 

Water-Supply Papers. — No. 219. Ground Waters and Irri- 
gation Enterprises in the Foothill Belt, Southern California ; by 
Walter C. Mendenhall. Pp. 180, 9 plates, 16 figures. 

No. 220. — Geology and Water Resources of a Portion of South- 
Central Oregon ; by Gerald A. Waring. Pp. 86, 10 plates, 1 
figure. 

No. 222. Preliminary Report on the Ground Waters of San 
Joaquin Valley, California ; by Walter C. Mendenhall. Pp. 
52, 1 plate. 

2. Canada : Geological Survey. A. P. Low, Director, 
Ottawa. — Recent publications from the Geological Survey of 
Canada, including the Department of Mines, are included in the 
following list : (See p. 239, vol. xxvi.) 

Seventeen maps, giving plans and sections of the Gold Districts 
of Nova Scotia. 

Preliminary Report on a Part of the Similkameen District, 
British Columbia; by Charles Camsell. Pp. 41, with folding 
map. 

Department of Mines, R. W. Brock, Acting Director. 

Report on a Portion of Conrad and Whitehorse Mining Dis- 
tricts, Yukon ; by D. D. Cairnes. Pp. 38, 8 plates, folding map. 

Preliminary Report on a Portion of the main Coast of British 
Columbia and adjacent Islands included in New Westminster and 
Nanaimo Districts; by O. E. Leroy. Pp. 56, 4 plates, 6 figures, 
folding map. 

Report on a Recent Discovery of Gold near Lake Megantic, 
Quebec ;• by John A. Dresser. Pp. 13, folding map. 

Report on the Landslide at Notre-Dame de la Salette, Lieore 
River, Quebec ; by R. W. Ells. Pp. 10, 7 plates. 

3. North Carolina Geological and Economic Survey, Joseph 
Hyde Pratt, State Geologist. 

Economic Paper No. 14. The Mining Industry in North Caro- 
lina during 1906 ; by J. H. Pratt. Pp. 142, 20 plates, 5 figures. 
Raleigh, 1907. 

Bulletin No. 16. Shade Trees for North Carolina ; by W. W. 
Ashe. Pp. 72, 10 plates, 18 figures. No. 17. Terracing of Farm 
Lands ; by W. W. Ashe. Pp. 72, 6 plates, 2 figures. Raleigh, 
1908. — The total value of the mineral productions of North 
Carolina, in 1906, was some $3,000,000, having increased from 
$1,800,000 since 1901. The most important items, making up 
two-thirds of the whole, are clay and coal. Following these are 



88 Scientific Intelligence. 

mica ($218,000), copper ($136,000), gold ($122,000) and monazite 
($126,000). The present report, by Dr. Pratt, gives an account 
of the mining industries of the state, particularly as related to 
gold, mica and monazite. 

The accompanying bulletins discuss the introduction of shade 
trees in the cities of the state, and the methods of terracing farm 
lands. This latter subject is particularly important in a region 
where the prosperity of the community depends so largely upon 
the preservation of the natural soil, which, when left to natural 
processes, soon suffers destructive erosion and is permanently lost. 

4. Report of the State Geologist of Vermont for 1907-8. 
Pp. 302, with 59 plates. — In the first portion (pp. 1-57) of this 
volume, the state geologist, Professor G. H. Perkins, treats of 
the mineral resources of the state, of which marble forms the 
bulk. In 1906 marble worth nearly four and one-half million dol- 
lars was produced, and of granite about four million dollars. T. 
Kelson Dale describes the Granites of Vermont ; H. E. Mervin, 
Some Late Wisconsin Shore Lines ; C. H. Hitchcock, the Geolog}^ 
of the Hanover, New Hampshire, Quadrangle ; G. H. Perkins, 
the Geology of Franklin and Chittenden Counties ; George E. 
Edson, the Geology of the Town of Swanton ; andC. H. Richard- 
son, the Geology of Newport, Troy and Coventry. 

An interesting discovery is the finding at St. Albans of a 
Middle Cambrian fauna with Paradoxides. The fine skeleton of 
a Pleistocene whale, Delphinapterus vermontamts, long shown 
at the State Museum, is here described and illustrated at great 
length by the state geologist. Professor Richardson seems to 
have fossil evidence indicating that the highly metamorphosed 
sedimentaries lying between the eastern foothills of the Green 
Mountains and the valley of Lake Memphremagog are of Cambro- 
Ordovician age. The fossils appear to be crushed graptolites. 

c. s. 

5. Thirty -second Annual Report, Indiana Department of 
Geology, W. S. Blatchley, State Geologist. Pp. 1231, with 22 
maps and 55 plates. Indianapolis, 1908. — Nearly 300 pages are 
devoted to a description of the soils of 17 counties of Southern 
Indiana by Messrs. Shannon, Lyons, Snider, Ward and Ellis. 
The vast oolitic limestone industrjr is described at length by the 
state geologist and associates. There are also the annual reports 
of the inspector of mines and natural gas. The chapter on the 
petroleum industry in 1907 states that since 1891 there have been 
sunk for oil 24,297 wells into the Trenton limestone and of these 
15,210 were producing last January. Last year the total shipped 
output was nearly five million barrels of 42 gallons each, bring- 
ing about 88 cents per barrel. 

About one-half of the book is devoted to " The Stratigraphy 
and Paleontology of the Cincinnati Series of Indiana," by Prof. 
R. E. Cumings. Sixty-seven local sections are described in detail 
and the fossils listed for each bed. This is followed by a general 
discussion of the Cincinnatian series. The greater part of the 



Geology. 89 

work is devoted to redescriptions of the 422 species collected, 
most of which are illustrated on 55 plates. The final paper by 
D. Reddick describes the mushrooms of Indiana. c. s. 

6. Illinois State Geological Survey, Year-Book for 1907 ; 
H. Foster Bain, Director. Bulletin No. 8, 374 pp., 23 pis., 32 
figs, with map. Urbana, 1907. — A number of papers contained 
in this report have been already published (this Journal, xxiv, 
447, xxv, 353-354, xxvi, 166). The remaining papers deal chiefly 
with economic subjects, including cement materials, clay industries, 
petroleum, artesian wells, lead and zinc, concrete, land reclama- 
tion, and valuable studies relating to the alteration of coal, with 
detailed investigations in certain areas. h. e. g. 

7. New Zealand Geological Survey, J. M. Bell Director. The 
Geology of the Cromwell Subdivision, Western Otago Division ; 
by James Park. Bull. No. 5 (new series). Pp. 86, pis. 20, maps 

10, geol. sections 6. Wellington, New Zealand, 1908. — The geo- 
logical section of the Cromwell area includes, in the Paleozoic sys- 
tem, two series : the Maniototo series of mica and chlorite schists 
which have an extreme thickness of 30,000 feet and a remarkably 
uniform schistosity, sills and dikes being almost completely absent 
over the wide area of five survey districts ; and the Kakanui series 
of upper schists which are less metamorphosed. The Pliocene is 
represented by the Manuherikia series of shales and clays, con- 
taining valuable seams of lignite. The region has been glaciated, 
as indicated by the moraines and extensive terraces and river 
gravels. All the glacial and fluviatile deposits are gold-bearing, 
and the principles underlying the concentration of gold in these 
gravels are discussed. Petrographic descriptions are given of the 
following rocks: Chlorite and mica-schist, greywacke, serpentine, 
hypersthene-diorite, feldspar-porphyrite, mica-gneiss, biotite- 
granite, augite-hypersthene-diorite, augite-diorite, hornblende- 
schist, hornblende-camptonite, sandstone, and chemical analyses 
are given of altered greywacke, serpentine, mica-schist, and 
chlorite-schists. Physiographically, the region is a part of the 
central Otago peneplain, now deeply dissected into " high table- 
top mountain ranges intersected by deep water courses and 
separated from each other by river valleys or cleft in twain by 
profound gorges". Mr. Park enters into an extensive discussion 
of the origin and development of the block mountains, with their 
intervening basins. h. e. g. 

8. Report on the Eruptions of the Soufriere in St. Vincent in 
1902, and on a Visit to Montagne Pelee in Martinique. Part 

11. The Changes in the Districts and the Subsequent History of 
the Volcanoes ; by Tempest Anderson. Petro graphical Notes 
on the Products of the Eruptions of May, 1902, at the Soufriere 
in St. Vincent; by John S. Flett. Phil. Trans. Roy. Soc. 
London, series A, vol. ccviii, pp. 275-332, 27 pis. London, 1908. 
— Dr. Anderson revisited St. Vincent and Martinique in 1907. 
His descriptions of the changes which have taken place between 
1902 and 1907 constitute an interesting study of the secondary 



90 Scientific Intelligence. 

phases of volcanic activity and also of the rapidity of erosion in 
volcanic materials, and the ease with which vegetation in tropical 
countries takes possession of a region which is absolutely barren* 
The sixteen plates from well selected photographs are excellent, 
and when compared with those taken in 1902 from approximately 
the same localities, constitute a history of stream development, 
erosion, and changes in volcanic materials which is very striking 
indeed. The present volume also contains the "Petrographical 
notes on the products of the eruptions of May, 1902, at the 
Soufriere in St. Vincent," by John S. Flett. The scientific world 
is fortunate in having these volcanoes studied by Anderson, La 
Croix, Hovey, and Heilprin, whose combined reports constitute 
perhaps the most elaborate treatise on any single volcanic disturb- 
ance in the world's history. The bibliography which accom- 
panies the report is fairly complete, but fails to mention the 
writings Of one of the most industrious students of this district, 
viz., Angelo Heilprin. h. e. g. 

9. The Geology and Ore deposits of the Coenr oVAlene Dis- 
trict, Idaho ; by Frederick Leslie Ransome and Frank Cath- 
cart Calkins. Professional paper 62. U. S. Geological Survey. 
Pp. 203, pis. xxix, figs. 23. Washington, D. C, 1908. — This 
report is of great interest to geologists since it embraces an area of 
404 square miles constituting the well known Coeur d'Alene min- 
ing district of northern Idaho and gives the detailed strati- 
graphic and structural geology of a portion of a region concern- 
ing which but little has been previously known. 

The district, as shown by the map, is one of maturely dissected 
mountainous topography showing a relief between river bottom 
and mountain top of about 4000 feet. It lies in the midst of a 
region which while not attaining elevations as great as certain 
others in the Cordillera, yet is, on the whole, of a particularly wild, 
rugged and forested character. The sedimentary rocks except 
for the surface gravels belong entirely to the great Algonkian 
system known as the Belt, from the earlier studies in the Belt 
mountains of Montana. The section here attains a thickness of 
17,200 feet, the base not exposed and the top removed by erosion. 
They vary from sandstones to argillites and throughout the 
greater portion of the system the argillaceous formations show 
marks of shallow water deposition and subaerial exposure. No 
great limestone formations such as the two which occur farther 
east are found in this region and it is concluded that the sedi- 
ments came from the west. A great fault, however, whose out- 
crop at the surface is known as the Purcell trench cuts off the 
system to the west, beyond this a highly metamorphic complex 
being exposed. The igneous rocks are all intrusive, the large 
masses varying from monzonites to syenites. The region is com- 
plexly folded and faulted but, as in other portions of this province, 
the rocks are remarkably free from regional metamorphism, giv- 
ing unusual opportunities for studying an Algonkian system. 

The mineral resources in the order of their present importance 
are (1) lead silver ores, (2) copper ores, (3) gold ores. j. b. 



Geology. 91 

10. Geologie der Steinkohlenlager ; by Dannenberg. Erster 
Teil. Pp. 197, 25 figs. (Gebriider Borntraeger.) Berlin, 1908.— 
This publication is evidently one of importance, and will be 
reviewed after the appearance of the second part, which the 
publishers announce will be issued sometime within the coming 
year. h. e. g. 

11. The Geology of Coal and Coal-mining ; by Walcot 
Gibson. Pp. x, 328, 8 plates. London, 1908 (Edward Arnold.) — 
Geologists and engineers will be interested in the series of works 
on economic geology, dealing with mining, quarrying, water 
supply, etc., to be issued under the general editorship of Pro- 
fessor Marr. The above volume is the first of the series and dis- 
cusses varieties of coal, their formation, origin, distribution, value 
of fossils in coal exploration, and methods for a study of exposed 
and concealed coal fields. Somewhat over half of the book is 
devoted to a critical discussion of the coal fields of the world, — 
those of Great Britain being treated with the greatest detail. 
Eighteen pages are devoted to the North American coal fields, 
and little use is made of the valuable investigations of the United 
States Geological Survey and the various state surveys. Mining- 
engineers, prospectors, and students of economic geology will 
find Dr. Gibson's book helpful reading. h. e. g. 

12. Physical History of the Earth in Outline ; by James B. 
Babbitt. Pp. 212, 6 illustrations. Boston, 1908 ( E. E. Sherman 
& Co.) — The object of this book is to explain the changes of cli- 
mate through geological time, on the assumption that the earth 
has a " proper rotary motion transverse to or across the diurnal 
rotation." An idea of the author's views in relation to orthodox 
science may be gained from the fact that he accepts Agassiz's 
account of glaciation in Brazil and believes the irregular coast 
lines of Maine, Norway, and the world in general, as well as the 
river system of North America, to be due to glaciation. h. e. g. 

13. Triassic Ichthyosaitria, toith special reference to American 
Forms ; by John C. Merriam. Memoirs of the University of 
California, Vol. i, No. 1. Pp. 155, text-figures 154, plates 1-18. 
Berkeley, 1908 (University Press). — The inaugural number of 
this new series of memoirs contains the results of twelve or more 
years of thorough work upon one of the most important groups 
of fossil vertebrates. Ten field-expeditions, sent out by the 
University of California and generously aided by a patron of that 
institution, have brought together a magnificent collection of 
ichthyosaurs from the marine Triassic of California and Nevada. 
While this collection forms the chief subject of the memoir, the 
European genera are also reviewed. Admirers of the author, 
and they are many, will be glad to learn that the present work, 
with all its breadth of treatment and refinement of detail, 
" can be considered as no more than a report of progress, as 
new material and additional information regarding the structure 
and affinities of the Triassic forms are constantly being obtained." 
In this case one of the great difficulties met with in paleonto- 



92 Scientific Intelligence. 

logical research has been tactfully overcome, viz., the question 
when an author should turn from the work of investigation and 
take the scientific public into his confidence. 

From the occurrence of considerably specialized ichthj^osaurs 
in the middle Triassic of regions as widely separated as Europe 
and western North America, it is argued that these forms had 
long been in existence as a marine tj^pe. Thus in the division of 
the memoir on general skeletal structure, the skull of Cymbo- 
spondylus, from the middle Triassic of West Humboldt Range in 
Nevada, is shown to be "the product of an ancestry which had 
expressed this special type of aquatic adaptation for a long 
period." In this connection it is interesting to learn that Cym- 
bosjjondylus is still the only Triassic ichthyosaurian genus 
represented by a well-preserved skull. In the known Triassic 
forms the number of presacral vertebrae is proved to be generally 
larger than in the typical Ichthyosaurus. This is slightly dis- 
turbing to the view held by some paleontologists that the 
Phytosauria are ancestral to the Ichthyosauria; and there are, the 
author states, a number of other characters apparently indicating 
that these two groups have arisen independently of each other. 
The principal points of difference between the Triassic and the 
Jurassic ichthyosaurs are arranged in parallel columns so as to 
demonstrate that the characteristics of the earlier genera are 
practically all nearer to those of land or shore animals, while 
" the characteristics of the later genera take the direction of 
specialization toward an adapted fish-like form." Yet on the 
whole, this evolution from a semi-aquatic reptilian type to one 
better fitted for life in deep water has been extremely gradual, 
and the unity of the Ichthyosauria (here divided into the families 
Mixosauridae and Ichthyosauridae) can not be questioned. At 
precisely what geological period the "unknown crawling ances- 
tor" of the ichthyosaurs asserted its independence from the 
parent shore-type, PrOf. Merriam wisely does not state ; but he 
places this event not later than the early part of the Trias — 
possibly at an even earlier period. The text-figures and plates 
are excellent ; and a carefully prepared index adds greatly to the 
value of the work. g. f. e. 

14. Catalogue of the Type and Figured Specimens of Fossil 
Vertebrates in the American Museum of Natural History, Part 
I— Fishes ; by L. Hussakof, Bull. Amer. Mus. Nat. Hist., Yol. 
xxv, pp. 1-103, with one diagram and plates i-vi, June, 1908. — 
In this exhaustive catalogue not only are all of the primary and 
secondary types of fossil fishes in the American Museum enumer- 
ated, but the condition of the specimen, together with the literary 
reference of the original description, or descriptions, is given. In 
all 49 text-figures are printed, while a number of specimens, mainly 
the types of Cope and Newberry, are illustrated in the excellent 
plates. The diagram gives in tabular form the classification and 
geological summary of the 562 types included in the catalogue, 

R. S. L. 



Geology. 93 

15. TJie Conard Fissure, a Pleistocene Bone Deposit in North- 
ern Arkansas, with descriptions of two new genera and twenty 
new species of Mammals ; by Barnum Brown. Memoirs Amer. 
Mus. Nat. Hist., Vol. ix, Part iv, pp. 155-208 ; pis. xiv-xxv. — Mr. 
Brown has made a most notable contribution to our knowledge of 
North American Pleistocene forms, not alone by this excellent 
work, but by the rare skill with which he explored and collected 
the material upon which the volume is based. The assemblage of 
animals contains thirty-seven genera and fifty-one species, of 
which four genera and twenty-one species are considered extinct. 
There is a notable absence of ground sloths, tapirs and proboscid- 
ians ; the fauna being characterized by northern forms, such as 
the musk ox and wapiti deer, among others. The condition 
of the bones, association and predominance of certain forms 
indicate that this fissure was the home of several contempora- 
neous species which preyed on others and brought their remains 
into it, the fauna being typically that of a forest region with open 
glades similar to present conditions. r. s. l. 

18. A Four-homed .Pelycosaurian from the Permian of Texas; 
by W. D. Matthew. Bull. Amer. Mus. Nat. History, Vol. xxiv, 
Art. xi, pp. 183-185. — This remarkable genus, Tetraceratops, is 
based upon the partially complete skull of a highly specialized, 
predaceous type. It is characterized mainly by the development 
of horn-like prominences — one in advance of each orbit and one 
arising from each of the premaxillaries. Among reptiles the 
nearest approach to them is seen in the carnivorous dinosaurs 
Ceratosaurus and Allosaurus, in both of which bony horns are 
found upon the prefrontals while the first-named genus bears a 
median nasal horn as well. The horns are more highly developed 
however, in Tetraceratops. Dr. Matthew's choice of name may 
lead to confusion, as the group of great horned Dinosauria, the 
Ceratopsia, includes the genera Ceratops, Diceratops and Tricera- 
tops, and the unrelated Tetraceratops should logically belong to 
the same group, if one were to judge from the name. r. s. l. 

11. Osteology of Blastomeryx and Phylogeny of the American 
Cervidce ; by W. D. Matthew. Bull. Amer. Mus. Nat. Hist., 
Vol. xxiv, Art. xxvii, pp. 535-562. — The lower Miocene fauna has 
supplied the connecting link between the Oligocene Leptomeryx 
and the later American deer in the form of Blastomeryx, the 
osteology of which is completely known. This enables Dr. 
Matthew to trace the evolution and relationships of the American 
Cervidse in a way which places the phylum on a plane with those 
of the Horses and Camels. Figure 14 is a diagram showing the 
evolution and migration into South America of the North 
American deer, while fig. 15 gives us a most valuable summary 
of the geological distribution and phylogeny of the American 
ruminants as a whole. r. s. l. 

18. Rhinoceroses from the Oligocene and Miocene deposits of 
North Dakota and Montana; by Earl Douglass. Ann. Car- 
negie Museum, Vol. iv, Nos. iii and iv, 256-266, pis. lxiii, lxiv. — 



94 Scientific Intelligence. 

In this brief paper Mr. Douglass describes a new species of the 
genus Aphelops, A. montanus, from the upper Miocene of Flint 
Creek Valley, Montana. Aphelops ceratorhinus, previously 
described by the writer, is made the subject of a more elaborate 
description, based upon the fully prepared type and upon other 
material referable to the same species. r. s. l. 

1 9. Fossil Horses from North Dakota and Montana y by Earl 
Douglass. Ibid, pp. 267-277, pis. lxv-lxviii — The middle Miocene 
of Montana has yielded a new genus and species of horse, Altippus 
taxus, while from the Loup Fork of Montana is described a new 
species of Merychippus, M. ynissouriensis. In the Oligocene Mr. 
Douglass found four species of Mesohippus, one of which he con- 
siders new, Mesohippus portentus. r. s. l. 

20. Some Oligocene Lizards ; by Earl Douglass. Ibid, pp. 
278-285. — In this paper Mr. Douglass describes a curious armored 
lizard from the Titanotherium beds of Montana and which he 
refers with a query to Glyptosaurus Marsh. The writer also 
enumerates the eight species of Glyptosaurus described by Pro- 
fessor Marsh, with brief diagnoses, for comparison with his 
present form. There are also described and figured skulls from 
the White River formation of Sioux Co., Neb., which the writer 
refers to Rhineura hatcheri Baur and Peltosauras granulosus ? 
Cope. r. s. l. 

21. Preliminary JSTotes on Some American Chalicotheres / by 
O. A. Peterson. American Naturalist, Vol. xli, pp. 733-752. — 
Mr. Peterson has been fortunate in securing from the remarkable 
Agate Spring Quarry (Lower Harrison) of Sioux Co., Nebr., a 
good deal of material which decidedly increases our knowledge 
of the curious genus, Moropus, first described by Professor Marsh. 
The present writer figures a partially complete skull, which he 
refers to Moropus elatus ? Marsh, as well as excellent photo- 
graphs of the fore and bind feet. Mr. Peterson's conclusions are 
thus summed up : "(1) That Moropusis, excepting its unguiculate 
feet, essentially a perissodactyl in structure. (2) That the later- 
ally compressed and cleft condition of the terminal phalanges is 
quite distinct in some of the early Perissodactyla, and that by 
adaptation through geological ages the unguals, as well as other 
parts of Moropus, were specially modified and should not, in the 
mind of the writer, be regarded as of ordinal importance. (3) 
That Moropus is generically separable from other known forms 
of the Chalicotherioidea." r. s. l. 



III. Botany and Zoology. 

1. The Harvard Botanical Station in Cuba. — In December 
1899, the writer, in company with Mr. Oakes Ames, Assistant- 
Director of the Botanic Garden of Harvard University, visited 
the sugar estate of E. F. Atkins, Esq., with the view of ascertain- 
ing what opportunity, if any, existed for conducting experiments 



Botany and Zoology. 95 

on the improvement of the cane. The estate is situated a short 
distance from the harbor of Cienfuegos, on the coast of the 
south-central part of Cuba. It was found that the estate, of 
large size, was supplied with all modern appliances for the manu- 
facture of sugar to the best advantage, and that facilities were 
there presented for conducting investigations in regard to cane- 
fertilization under favorable conditions. The services of the skilled 
chemists at the factory on the estate were placed at our disposal, 
for the determination of many questions arising with reference 
to yield and sugar-content of the cane. The results of the pre- 
liminary examination were so favorable that the work of crossing 
was actively begun, and continued the next year. Mr. Robert 
M. Grey, who had been very successful in hybridizing orchids, 
was invited to make a careful examination of the conditions pre- 
sented by the locality, and he gave a conservative report which 
was decidedly encouraging. In 1902 we had also an inspection 
of the place by Mr. C. G. Pringle, who is well known as a hybrid- 
izer of cereals, and he likewise expressed the opinion that the 
estate afforded good opportunities for fruitful crossing of varie- 
ties. In the early part of the next year, Mr. Grey was made super- 
intendent of the newly formed station, and began systematic 
experimenting along the lines laid down by the sugar-cane 
experts in Java. We were so fortunate as to have, also, the coun- 
sel of Dr. John C. Willis, Director of the Royal Botanic Gardens 
of Ceylon, who made a careful investigation of the capabilities 
of the station. His favorable report led to immediate extension 
of the original plan, and the development has steadily progressed 
with practially no interruptions except during the short period of 
the late insurrection. At present, ten years after the first work 
in crossing was done at this locality, the experiment may be 
regarded as successful. At great cost we have secured from the 
most remote localities, authentic specimens of the finest canes now 
known, and the varieties have been carefully perpetuated under 
the best conditions for each. Meanwhile we have added to the 
station the principal economic plants of the warm and the hot 
tropics, and have indicated the lines of research likely to prove 
most satisfactory. The results which have been reached by Mr. 
Grey have depended upon his skill both as a cultivator and as 
a plant-breeder. His monthly reports exhibit a very wide range 
of experimenting and wholly fruitful outcome. 

Fortunately, from the very first we have been able to maintain 
agreeable relations with sister stations in Cuba and the West 
India Islands, and have received constant #id from all officials. 
Of course by reciprocity our results are placed at their disposal. 
Some of the more striking of the results have been published in 
the Boletin oficial (Cuba) and in the West India Bulletin. The 
monthly reports from the station are now becoming of so much 
interest that some of the particular features seem to require pre- 
sentation in a regular publication. As to the form of this, no decis- 
ion has yet been made. 



96 Scientific Intelligence. 

Mr. Atkins, who has sustained the enterprise from the outset, 
expresses himself as satisfied with the substantial results reached. 
It is a great pleasure for the director, assistant-director, and 
head-gardener of the Harvard Botanic Garden to aid the rapid 
development of the station by every means in their power. 

G. L. G. 

2. Handbuch der Bluten-biologie, von Dr. Paul Knuth. 
Leipzig (Wilhelm Englemann). Handbook of Flower- Pollina- 
tion, by Dr. Paul Knuth ; translated by J. R. Ainsworth Davis, 
M.A. Oxford (Clarendon Press). — This excellent translation 
brings into two volumes the subject matter contained in the three 
German volumes (bound in five parts). The original work is based 
on the interesting treatise by Hermann Milller which, it will be 
remembered, greatly stimulated research in the attractive field 
of the pollination of flowers. The innumerable contributions to 
this subject soon outgrew the limits which could naturally be 
assigned to a second edition, and necessitated entire reconstruction. 
Professor Knuth undertook this reconstruction, while carrying 
the pressure of the double burden of deep affliction and impaired 
health. It was hoped that a long journey, which he prosecuted 
in the search for fresh material, might lighten this burden, but 
this did not prove to be the case. His early death left an 
immense mass of material partly published, but to a great extent 
uncoordinated. Willing hands have from this material completed 
a lasting monument to the author. Its comprehensiveness and 
accuracy will enable this work long to maintain its place as a 
memorial. The translation presents the whole treatise in a very 
convenient form for the English-speaking student. g. l. g. 

3. A Convenient Clearing and Mounting Agent. — To a 
perfectly clear solution of potassium silicate add one third to one 
half volume of glycerine and after warming the whole, slightly 
shake until the two liquids are thoroughly mixed. The result- 
ant clear liquid has proved to be an efficient agent for 
clearing such objects as pharmaceutical powders and the like. 
If the powder, for instance, capsicum, is placed on a glass slide 
and the glycerin-silicate placed carefully thereon, a cover-glass 
can be at once put in position and held for a couple of minutes or 
so, when it will be found to have "set" more or less firmly, constitut- 
ing a fairly good mount for use within the next few days. If the 
edge of the cover be carefully cemented by a mixture of potas- 
sium silicate and barium sulphate, the mount becomes permanent. 
Two precautions must be observed ; first, use enough of the 
glycerin-silicate to flow, at the outset, to the edge of the cover- 
glass, all around, and second, on no account allow any of the 
agent to touch the front of the objective. 

A mixture of sodium silicate and glycerin (see Schurhoft, 
quoted in Journal of Royal Microscopical Society, 1902, p. 622) 
does not appear to clear the objects under inspection quite so 
well as the one here suggested. g. l. g. 



Botany and Zoology. 97 

4. Economic Zoology, An Introductory Text-booh in Zoo- 
logy, with special reference to its applications in Agriculture, 
Commerce, and Medicine ; by Herbert Osborn. Pp. xv, 490, 
with 269 figures. New York, 1908 (The Macmillan Company). 
— A recent tendency to emphasize those features of a science 
which have a practical hearing on human affairs finds expression 
in this little text-hook, which differs from others of its class mainly 
in the relative amount of attention devoted to those animals 
which are of economic importance. The book has the usual sys- 
tematic arrangement, and all the groups of the animal kingdom 
are included, but where a choice of examples is possible the 
species which concerns the human welfare is selected and its 
significance discussed. The numerous illustrations are largely 
selected from other books issued by the same publishing house, 
and are in many cases not so well printed as in the works from 
which they are taken. w. r. c. 

5. A Text-booh of the Principles of Animal Histology ; by 
Ulric Daiilgren and William A. Kepner. Pp. xiii, 515. New 
York 1908 ( The Macmillan Company). — In marked contrast to 
practically all the text-books of histology in the English language, 
which deal largely or exclusively with human or mammalian struc- 
tures, this new book discusses the tissues of all classes of animals. 
It is therefore possible to treat the subject much more broadly and 
satisfactorily than has hitherto been done. The increased range of 
material makes possible the incorporation of new and vastly im- 
proved illustrations in place of the worn-out cuts of most other his- 
tologies. The animal tissues are discussed in systematic order, 
and the modifications of the tissues of the invertebrates find their 
place beside the better known vertebrate structures. The orig- 
inality of the work is most praiseworthy, and although some of 
the illustrations are rather crude, they give at least a diagram- 
matic illustration of the structures they are intended to represent, 
and will be welcomed by all teachers of the subject. This book 
will doubtless supplant in many cases the older works, adapted 
chiefly to the needs of the student of medicine, which have 
hitherto alone been available for scientific college courses. 

W. R. C. 

6. Archiv fur Zellforschung ; hrsg. von Richard Gold- 
schmidt. Bd. ], Heft 1-4 ; pp. 622, pi. i-xxi. Leipzig, 1908 (Wil- 
helm Englemann). — This newly established journal, devoted to the 
study of cellular structure and phenomena of animals and plants, 
takes its place at once among the highest type of biological publica- 
tions. It is intended for the publication of the results of original 
research, and is illustrated by the finest lithographic plates. 
This first volume contains fifteen original contributions by some 
of the most prominent cytologists of Europe, the subjects of their 
investigations covering a wide range of cytological problems, 
both in plants and animals. w. r. c. 

7. Ueber die Eibildung bei der Milbe Pediculopsis graminum; 
von Exzio Reuter. — A reprint from the Festschrift fur Palmen 

Am. Jour. Sci.— Fourth Series, Vol. XXVII, No. 157.— January, 1909. 

7 



98 Scientific Intelligence. 

( Helsingfors, 1907 ) containing an account of a case of ovogenesis 
in which certain of the egg cells, possibly those normally destined 
to produce the male sex, are absorbed by the more vigorous 
(female ?) cell's, thus producing a preponderance of female indi- 
viduals, w. e. c. 



IV. Miscellaneous Scientific Intelligence. 

1. Artificial Daylight for Use loith the Microscope ; by Fred. 
Eugene Weight. (Communicated from the Geophysical Labora- 
tory.) — For microscopic work with rocks, and especially with arti- 
ficial products, daylight on dark winter days is a factor much to be 
reckoned with, and the want of a good substitute for it has long 
been felt by the writer. Many different kinds of artificial light 
are available, but the results obtained by their use have not, on 
the whole, been satisfactory. The natural colors of mineral sec- 
tions, and also interference colors, appear abnormal and unnatural 
in artificial light, the illumination of the field is not even, and 
good interference figures on small plates are not as a rule obtain- 
able. Recently, however, the following illuminating device has 
been tried and has proved so satisfactory that it deserves brief 
mention. 

The source of light is an acetylene burner fed by a J. B. Colt 
generator No. 102, and placed at the focus of a large condens- 
ing lens of about 20 cm focal length. Compared with daylight, 
acetylene light is slightly too strong in yellow, bat by passing the 
parallel rays from the condenser lens through a pale blue cobalt 
glass plate of the proper intensity, this difficulty is eliminated 
and the field appears white and the colors, both natural and inter- 
ference colors, are normal and correct. In certain tones, very 
slight differences can be -detected, but these are of an order 
noticeable in different sections of the same substance and are not 
serious. With this arrangement, acetylene light, condenser lens 
and pale blue glass ray filter, the observer has at hand a source of 
illumination which is practically identical in its effect with day- 
light, and which is available at any moment. 

2. Ion. A Journal of Ulectrotonics, Atomisties, lonology, 
Radio-activity and Raumchemistry. Edited by Charles H. 
Walter* Vol. I, Fasc. 1. Pp. 80. November, 1908. London 
(16 Heathfield Gardens, Turnham Green, W.). — An unavoidable 
result of the recent development of science along special lines has 
been the establishment of many new journals of more and more 
restricted field. This multiplication of periodicals is in many 
respects an admirable thing, having a decidedly stimulating effect 
upon the workers immediately interested, but, at the same time, 
it has its drawbacks and adds a somewhat heavy burden upon 
science in general. In Germany, in particular, these specialist 

* The name of Frederick Soddy appears on the title page as an editor-in- 
chief, but in Nature for Nov. 26, Professor Soddy announces that he has 
"withdrawn from all connection with the journal." 



Miscellaneous Intelligence. 99 

periodicals are very numerous ; in England there have been but 
few. There is more room, therefore, for this new journal, the 
"Ion," which is started with the design of covering the field of 
physico-chemistry, or in other words the various special branches of 
physics in which the electron is the essential conception ; one of 
the most highly developed of these is the subject of radio-activity. 
Associated with the editor-in-chief are the following : Sv. Arrhe- 
nius, Stockholm ; W. H. Bragg, Adelaide ; A. S. Eve, Montreal ; 
O. Hahn, Berlin ; W. H. Julius, Utrecht ; A. Werner, Zurich ; G. 
Bruni, Padua ; Mde. Curie, Paris ; Guilleaume, Paris ; Van't 
Hoff, Berlin ; W. Marckwald, Berlin ; W. Wien, Wiirzburg. 

The first number of Ion bears the date of November, 1908, and 
opens with an account of M. Henri Becquerel, the founder of 
radio-activity, by F. Soddy, who also contributes a paper on the 
charge carried by the a-particles. Other articles are on uranium 
and geology by John Joly ; the transmission of energy in the 
world of electrons by H. W. Julius ; and actinium C by Otto 
Hahn and Lise Meitner. Following these articles are some 
twenty pages devoted to reports of papers published elsewhere 
and notices of new books. 

The new journal will be issued in monthly numbers of 64 to 
80 pages, at the price of 30s. per volume ; the editorial address is 
given above. 

3. Life and Letters of Herbert Spencer / by David Duncan. 
Two volumes. Pp. viii, 414 and vii, 444, with seven full-page 
illustrations. New York, 1908 (D. Appleton & Company). 
— These two attractive volumes portray the human side of the 
life of the great philosopher and form a fitting complement to 
Spencer's own voluminous autobiography. The author and 
compiler was commissioned by Spencer himself to prepare such 
a work for publication, and to incorporate certain unpublished 
papers and portraits. From the vast amount of correspondence 
which passed between Spencer and members of his family, friends, 
and the scientific men of the day, the author has made such 
judicious selection, and has woven the whole together so skill- 
fully by explanation and anecdote, that the work can be read with 
perfect smoothness. And at the same time the reader gains such 
vivid pictures of the real man, his attitude toward the great 
problems of his day, and with frequent glimpses of his scientific 
associates, as could hardly be secured in any other manner. The 
letters are arranged largely with refereuce to subject matter, but 
with some regard for chronological order, and the work is illus- 
trated by a number of portraits not only of Spencer at successive 
periods of his life, but also of persons with whom he corresponded. 
In five appendices appear several hitherto unpublished essays, 
together with lists of his writings (the titles covering fifteen 
pages) and the honors which were offered the great scholar by 
upwards of thirty universities and learned societies. It should 
be added that these honors were, however, with few exceptions 
declined. - w. n. c. 



100 Scientific Intelligence. 

4. American Association for the Advancement of Science. — 
The sixtieth meeting of the American Association is held at 
Baltimore, under the auspices of Johns Hopkins University, dur- 
ing the week from Dec. 28, 1908 to Jan. 2, 1909; Prof. T. C. 
Chamberlin of Chicago is President. This is the seventh of 
the " Convocation week" meetings ; some twenty-four affiliated 
societies meet at Baltimore at this time. A Darwin commemo- 
rative meeting takes place on January 1, under the combined 
auspices of the Association and the Society of Naturalists. Also 
a meeting of the American Health League is called for Dec. 31, 
in connection with Section I of the Association, as a symposium 
on " Public Health." 

5. The Nature of Enzyme Action; by W. M. Bayliss. Pp. 
90. London, 1908 (Longmans, Green & Co.). — This is the first of 
a series of monographs on bio-chemistry planned to supplement 
the current text-book treatment of topics in rapidly developing 
departments of this science. Enzyme action is shown to be a type 
of catalytic reaction the features of which are subjected to critical 
analysis by the author, and compared with the behavior of other 
catalysts. Special consideration is devoted to the nature of 
colloids as exemplified in enzymes, to the reversibility of enzyme 
action, and to changes in the rate of reactions as affected by them. 
Other appropriate details, such as the relation of the enzyme to 
its substrate, the influence of "co-enzymes" and " anti-enzymes," 
temperature and concentration, etc., are also discussed. The 
treatment is quite original. l. b. m. 

6. Hivista cli Scienza. — This valuable " International Review 
of Scientific Synthesis," commenced in 1907, has now completed 
its fourth volume. A recent number includes eight articles, 
among which may be mentioned the following : G. H. Bryan on 
the diffusion and dissipation of energy ; W. Ritz on the role of 
the ether in Physics ; G. Haberlandt on motion and sensation in 
the plant world ; G. Schiaparelli, on the astronomy of the 
Bablyonians. 

B. G. Teubner's Verlag auf deni Gebiete der JVlathematik Naturwissen- 
schaften and Technick nebst Grenzwissenscliaften. Mit einem Gedenktage- 
buche fur Mathematiker und den Bildmissen von G. Galilei, etc. Dem IV. 
Internationalen Mathematiker-Kongress in Eom. 6-11. April 1908. Pp.392. 

Obituary 

Dr. Oliver Wolcott Gibbs, from 1863 to 1887 Rum ford 
Professor of Applied Science in Harvard University and for 
thirty years an Associate Editor of this Journal, died at his home 
in Newport, R. I., on December 9, in his eighty-seventh year. 
A notice is deferred until a later number. 

Professor William Edward Ayrton, the eminent English 
engineer, died on November 8, at the age of sixty-one years. 



New Circulars. 



84 : Eighth Mineral List : A descriptive list of new arrivals, 

rare and showy minerals. 

85 : Minerals for Sale by Weight: Price list of minerals for 

blowpipe and laboratory work. 

86 : Minerals and Rocks for Working Collections : List of 
common minerals and rocks for study specimens ; prices 
from i% cents up. 

Catalogue 26: Biological Supplies : New illustrated price list 
of material for dissection ; study and display specimens ; 
special dissections; models, etc. Sixth edition. 

Any or all of the above lists will be sent free on request. We are 
constantly acquiring new material and publishing new lists. It pays to 
be on our mailing list. 



Ward's Natural Science Establishment 

76-104 College Aye., Rochester, N. Y. 

Ward's Natural Science Establishment 

A Supply-House for Scientific Material. 

Founded 1862. Incorporated 1890. 

DEPARTMENTS : 

Geology, including Phenomenal and Physiographic. 
Mineralogy, including also Kocks, Meteorites, etc. 
Palaeontology. Archaeology and Ethnology. 

Invertebrates, including Biology, Conchology, etc. 
Zoology, including Osteology and Taxidermy. 
Human Anatomy, including Craniology, Odontology, etc. 
Models, Plaster Casts and Wall-Charts in all departments. 



Circulars in any department free on request ; address 

Ward's Natural Science Establishment. 

76-104 College Ave., Rochester, New York, TJ. S. A. 



CONTENTS. 

Page 
Art. I. — Diopside and its Relations to Calcium and Magne- 
sium Metasilicates ; by E. T. Allen and VV. P. White. 
With Optical Study ; by F. E. Wright and E. S. 

Larsen. (With Plate I) 1 

II. — California Earthquake of 1906; by G. K. Gilbert 48 

III. — Descriptions of Tertiary Insects ; by T. D. A. Cock- 

erell. PartV _ __. 53 

IV. — Electrolytic Estimation of Lead and of Manganese by 
the Use of the Filtering Crucible ; by F. A. Gooch and 

F. B. Beyer . . , 59 

V. — Specific Radio-Activity of Thorium and its Products ; 

by G. C. Ashman _. 65 

VI. — Coronas with Mercury Light ; by C. Barus 73 

SCIENTIFIC INTELLIGENCE. 

Chemistry and Ph ysics — Attempt to Produce a Compound of Argon, Fischer 
and Iliovici : Explosive Crystallization, Weston, 82. — Constituents of 
Ytterbium, A. v. Welsbach : New Form of " Tin Infection," vow Hass- 
linger : Heat Evolved by Radium, von Schweidleb and Hess, 83. — 
Positive Rays, "W. Wien : Spectral Intensity of Canal Rays, J. Stark and 
W. Stenberg : Canal Rays, J. Stark : Potential Measurements in the 
dark Cathode Space, W. Westphal, 84. — Elements of Physics, Nichols 
and Franklin : Text -Book of Physics, 85. 

Geology — Publications of the United States Geological Survey, 86. — Canada 
Geological Survey : North Carolina Geological and Economic Survey, 
87.— Report of the State Geologist of Vermont for 1907-8 : Thirty- 
second Annual Report, Indiana Department of Geology, 88. — Illinois 
State Geological Survey, Year-Book for 1907 : New Zealand Geolog- 
ical Survey : Report on the Eruptions of the Soufriere in St. Vincent in 
1902, and on a Visit to Montagne Pelee in Martinique, T. Anderson and 
J. S. Flett, 89. — Geology and Ore deposits of the Coeur d' Alene District, 
Idaho, 90.— Geologie der Steinkohlenlager, Dannenberg : Geology of 
Coal and Coal-mining, W. Gibson : Physical History of the Earth in Out- 
line, J. B. Babbitt : Triassic Ichthyosauria, with special reference to 
American Forms, J. C. Merriam, 91. — Fossil Vertebrates in the American 
Museum of Natural History, Part I — Fishes, L. Hussakof, .92.— Conrad 
Fissure, B. Brown : Four-horned Pelycosaurian from the Permian of 
Texas, W. D. Matthew : Osteology of Blastomeryx and Phylogeny of the 
American Cervidge, W. D. Matthew : Rhinoceroses from the Oligocene 
and Miocene deposits of North Dakota and Montana, E. Douglass, 93. — 
Fossil Horses from North Dakota and Montana, E. Douglass : Some Oli- 
gocene Lizards, E. Douglass : Preliminary Notes on Some American 
Chalicotheres, 0. A. Peterson, 94. 

Botany and Zoology — Harvard Botanical Station in Cuba, 94. — Handbuch 
der Bliiten-biologie, P. Knuth : Convenient Clearing and Mounting Agent, 
96. — Economic Zoology : Text-book of the Principles of Animal Histology, 
U. Dahlgren and W. A. Kepner : Archiv. fur Zellforschung, R. Gold- 
schmidt : Ueber die Eibildung bei der Milbe Pediculopsis graminum, 97. 

Miscellaneous Scientific Intelligence— Artificial Daylight for Use with the 
Microscope, F. E. Wright : Ion ; A Journal of Electrotonics, Atomistics, 
Ionology, Radio-activity and Raumchemistry, 98. — Life and Letters of 
Herbert Spencer, D. Duncan, 99. — American Association for the Advance- 
ment of Science : Nature of Enzyme Action : Rivista di Scienza, 100. 

Obituary— O. W. Gibbs : W. E. Ayrton, 100. 



ur. uyru^\oTe^ 

Librarian U. S. Nat. Museum. 

VOL. XXVII. • FEBRUARY, 1909. 



Established by BENJAMIN SIIIIMAN in 1818. 
THE 

AMERICAN 

JOURNAL OF SCIENCE. 

Editor: EDWARD S. DANA. 

ASSOCIATE EDITORS 

Professors GEOKGE L. GOODALE, JOHN TEOWBKIDGE, 
W. G. FAKLOW and WM. M. DAVIS, of Cambridge, 

Professors ADDISON E. VERRILL, HORACE L. WELLS, 
L. Y. PIRSSON and H. E. GREGORY, of New Haven, 

Professor GEORGE F. BARKER, of Philadelphia, 
Professor HENRY S. WILLIAMS, of Ithaca, 
Professor JOSEPH S. AMES, of Baltimore, 
Mr. J. S. DILLER, of Washington. 



FOURTH SERIES 

No. 158— FEBRUARY, 1909. 

YOL. XXVII— [WHOLE NUMBER, CLXXVIL] 

WITH PLATES II-IV. 

NEW HAVEN, CONNECTICUT. 

1909. 



THE TUTTLE, MOREHOUSE & TAYLOR CO., PRINTERS, 123 TEMPLE/STREET 



thsoi 



Published monthly. Six dollars per year, in advance. $6.40 to countries in the 
Postal Union ; $6.25 to Canada. Remittances should be made either by money orders, 
registered letters, or bank checks (preferably on New York banks). 



RARE CINNABARS FROM CHINA. 

During the past month crar correspondent in China ran across fine, 
remarkable specimens of these Cinnabars and we immediately ordered 
them secured for us, and they will reach us about the time this Journal is 
published. Write for illustrated pamphlet and particulars. 

NEW ARRIVALS. 

Euclase, Capo do Lane, Brazil ; Chalcocite, Bristol, Conn. ; Columbite, 
Portland, Conn.; Iolite, Guilford, Conn.; Monazite, large loose xls. and in 
matrix, Portland, Conn.; Uraninite, crystal in matrix, Portland, Conn.; 
Eglestonite, Terlingua, Texas ; Patronite, S. A. ; Benitoite, San Benito Co. , 
Cal.; Neptunite, Cal.; Crookesite, Hedenbergite, Sweden; Lievrite, Elba; 
Polybasite, Hungary and Durango, Mexico ; Josephinite, Oregon ; Herderite, 
Poland, Maine ; Smithsonite, Kelly, N. M. ; Calif oraite, Tulare Co., Cal. ; 
Cobaltite, loose crystals and in matrix. Cobalt, Ont., and Tunaberg, 
Sweden ; Apatite, Auburn, Maine ; Vivianite, large crystals, Colo. ; Vanadi- 
nite, Kelly, N. M. ; Olivenite, Utah ; Sartorite, Canton Wallis ; Jordanite, 
Binnenthal ; Mohawkite, Algodonite, Domeykite, Michigan ; Crocoite, Sibe- 
ria and Tasmania ; Cinnabar, Cal., Hungary and China; Dioptase, Siberia; 
Diopside, with Essonite, Ala, Piedmont ; Embolite, Australia ; Gypsum, 
twin crystals, Eisleben, Thuringia ; Diamond, in matrix, New Vaal River 
Mine, South Africa ; Argentite, Mexico ; Freiberg, Saxony ; Pyrargyrite, 
Saxony and Mexico ; Wulfenite on Aragonite, Organ Mts. N. Mexico ; 
Celestite, Sicily ; Pyromorphite, Ems, Germany, Phoenixville, Pa. ; Miller- 
ite, Antwerp, N. Y. ; Sylvanite, Colo.; Tellurium, Colo.; Tourmalines, new 
find, San Diego Co. , Cal. ; Tourmalines, beautiful sections from Brazil ; Bro- 
chantite on Chrysocolla, Utah ; Pink Beryl, small and large, Mesa Grande, 
Cal.; Kunzite, small and large, Pala, Cal; Sphene, Binnenthal; Titan! te, 
Tilly Foster, N. Y. ; Tetrahedrite, Utah and Hungary ; Realgar, Hungary ; 
Opal, Caribou River, Queensland ; Octahedrite, var. Wiserine, Binnenthal ; 
Heulandite, Iceland ; Torbernite, Eng. ; Bismuth, native, Cobalt, Ont. and 
Conn.; Silver, native, polished, Ont.; Emerald, loose and in matrix, Ural 
and Bogota ; Topaz crystals with rare planes, Ural ; Zircon crystals, loose, 
Ural ; Green and Cinnamon Garnets, Minot, Maine ; Vesuvianite, Poland, 
Maine, Italy and Tyrol ; Zeolites, beautiful specimens from Erie Tunnel, 
Patterson and Great Notch. 

CUT GEMS. 

' Garnets, green and red ; Aquamarines ; Zircons, all shades ; Sapphires, 
all shades ; Star Sapphires and Star Rubies ; Chrysoberyl, Cats-eye ; Spinels, 
all shades ; Topaz, pink, blue, brownish and golden color ; Pink Beryl ; 
Sphene; Tourmaline, all shades ; Amethyst, Siberia, royal purple color ; 
Andalusite ; Star Quartz ; Peridote ; Opal matrix, Mexico and Australia ; 
Precious Opal, Australia, Mexico and Hungary ; Hyacinth ; Turquoise, Mex- 
ico and Persia ; Kunzite ; Reconstructed Rubies and Sapphires ; Emeralds ; 
Opal Carvings, such as pansies, vine leaves with bunches of grapes, and 
other small Opal novelties ; Antique and Modern Cameos ; Antique, Mosaic 
and other semi-precious stones. 

Let us know your wants, and we will send the specimens on approval to 
you. Write for our new circular today. 

A. H. PETEREIT, 

81—83 Fulton Street, New York City. 



THE 



AMERICAN JOURNAL OF SCIENCE 

[FOURTH SERIES.] 



Art. TIL — Revision of the Protostegidce ; by G. R. 
Wieland. '(With Plates II-IY.) 

[Contributions from the Paleontological Laboratory of Yale University.] 

There is no family among all American fossil turtles which, 
following the discovery of its initial type, has so steadily 
yielded new forms and additions to our knowledge of the 
structure and history of marine turtles as the Protostegidse. 
True enough, no further members of the family were noted 
and few specimens were collected for twenty years after Cope's 
original discovery of Protostega gigas / but then came the 
addition of the related genus Archelon from the Pierre Cre- 
taceous in 1895, since which time scarcely a year has passed 
without yielding new data to the structure, extent and signifi- 
cance of the Protostegidse. 

Indeed, even before the discovery of Archelon the attention 
of the brilliant and incisive Baur had been turned to Protos- 
tega ; and since then Hay, Case, Williston and Wieland have 
all contributed in turn to the literature of the Protostegidse, — 
while in Europe Dollo has published papers of the greatest 
supplemental interest dealing with the origin of marine turtles. 

Furthermore the collection of the splendid cotypes of Pro- 
tostega gigas showing the complete limb structure, now in the 
Carnegie Museum of Pittsburg, and more recently the mount- 
ing for exhibition of the huge type of Archelon ischyros in 
the Yale Museum, have contributed much toward the increas- 
ingly accurate picture of the Protostegidse. With the descrip- 
tion of new species, meanwhile, and the appearance of the 
great volume of Hay — easily the foremost contribution to the 
literature of the Testudinata yet made — it is already evident 
that the Protostegidse include a series of forms of the greatest 
structural interest, and that further additions to the family are 

Am. Jour. Sci. — Fourth Series, Vol. XXVII, No. 158. — February, 1909. 

8 



102 G.R. Wieland— On Marine Turtles. 

certain to be made. Moreover, to all these newer facts and 
viewpoints we are enabled to add the description of a new 
species, calling for analysis of the gronp. 

To these forewords to the present revision I wish to add 
praise for the painstaking labor bestowed upon the mounting 
of Archelon by the Yale Museum preparator Mr. Hugh 
Gibb. Likewise we are indebted to the rare skill of the well 
known scientific illustrator and artist Mr. P. Weber for the 
illustrations following. 

CHELONIOIDEA Baur. 

SUPERFAMILT OF THE CrYPTODIRA. 

A parieto-squamosal arch ; palatine foramen and free nasals 
sometimes present (Desmatochelyidge); fourth cervical cyrtean, 
with the centra of the sixth to eighth less modified in Creta- 
ceous than in recent forms. 

The five great marine families, namely, the Cheloniidse, 
Protostegidse, Desmatochelyidse, Toxochelyidee, and Dermo- 
chelyidse, all doubtless independently acquired their equipment 
for life in the sea. ]5 

Family Protostegidce Cope. 

Turtles with highly specialized thalassic humeri, but with 
three or more claws. A leathery hide and osteodermal arma- 
ture evidently present. Carapace usually greatly reduced in 
later forms, the disk investing less than one-half the rib lengths. 
Plastron not markedly reduced. Peripherals serrate to 
strongly digitate on their interior borders ; intra-peripheral 
dermogene ossicles sometimes present (known in Archelon 
only). Plastron very large, dactylosternal, with prominent 
fontanelles ; epiplastra small, out-turned, separate, and wholly 
supported by the very large T-shaped entoplastron ; hyo- and 
hypoplastra moderately digitate (Protostega advena) to strongly 
digitate {Archelon) ; xiphiplastra short and bowed. Pelvis 
with obturator foramina enclosed by complete ischio-pubic 
border. Coracoid extending all the way back to the pre-pubis 
except possibly in P. Copei. Skull large ; temporal region 
broadly roofed over ; descending processes of parietals ; ant- 
orbital projection marked ; quadrato-squamosal vertex much 
depressed ; narial aperture more or less upturned ; choanae far 
forward, opening free behind vomer. 

Genus Protostega Cope, 1872. 1 
Premaxillary beak less developed than in Archelon / maxilla 
with rather broad grinding surface, which extends backward 
to behind front of orbit. Lower jaw with rami early coossi- 



G. R. Wieland—On Marine Turtles. 103 

fied. Neuralia normal so far as yet seen, and without median 
pits or grooves. . Radial process of humerus large and project- 
ing. 

Species of Protostega. 

A. Niobrara Cretaceous : — 

1. A medium-sized to large turtle, with a thin cara- 

pace investing one-third the rib lengths, and 
interior borders of marginals splitting into 
medium-sized digitations. P. gigas. 

2. A small turtle, with more than half the rib 

lengths expanded and with less reduced plas- 
tron than the preceding ; the hyo- but not the 
hypoplastic meeting on the median line ; 
xiphiplastrals only slightly bowed ; marginalia 
heavy and without digitation of interior 
borders. P. advena. 

3. A medium-sized turtle, with a comparatively 

thick carapace investing the proximal half of 
the ribs; plastral form nearly as in P. gigas 
but with more numerous digitations and smaller 
fontanels ; marginal borders serrate rather than 
smooth ; limb bones relatively short and small. 
This form has the heaviest shell of any Pro- 
tostegid. The carapace is little more, and the 
plastron less reduced than in the Cheloniidae. P. Copei. 

4. A large turtle, with xiphiplastra nearly joined 

on the median line and epiplastral pittings on 

outer anterior projection of hyoplastra. P. potens. 

B. Pierre Cretaceous : — 

5. An immense turtle, with neuralia like P. Copei, 

but humerus without a markedly strong radial 
process ; marginalia strongly digitate on inte- 
rior borders. P. (Archelon) Marshii. 

Genus Archelon Wieland, 1896. 6 

Premaxillary beak greatly developed and strongly decurvecl ; 
crushing surface of upper jaw set far forward and limited to 
vomero-maxillar region ; lower jaw with rami not coossified 
until old age. Neuralia greatly reduced, to partly abseut 
anteriorly, and replaced by epineuralia with a deep median 
sulcus nearly continuous to the eighth true [underlying] 
neural ; tenth rib relatively large, free, and extending out to 
marginalia. Radial process of humerus weak. 

Archelon ischyros from the uppermost Pierre Cretaceous of 
the valley of the South Fork of the Cheyenne River is the 
only species. 



104 G. E. Wieland — On Marine Turtles. 

Protostega Copei sp. nov. (Figures L-4). 

A new species, which may be appropriately named for the 
illustrious discoverer of the Protostegidae, is indicated by the 
most complete and best conserved specimen referable to its 
family, thus far obtained. This splendid fossil is from the 
Niobrara chalk of the Hackberry Creek Valley, Gove county, 
Kansas, and was found in the summer of 1905, by the veteran 
collector and explorer, Mr. Charles H. Sternberg. 

Fig. 1. 



«.,#*!; x:.— — -•* 






.■■£% 

' ■'&:?>■■ ■:■;■; 


-'- ■--' '" %;'4 





Figure I.— Protostega Copei. Photograph of skull and lower jaw of type 
as mounted in the Yale University Museum by bringing together the disso- 
ciated and for the greater part but little crushed cranial elements. One- 
fourth the natural size. 

Only minor portions of certain boundaries had to be restored. A little 
attention will at once reveal the limiting sutures of the premaxillary, max- 
illary, frontal, post-frontal, parietal, jugal, post-frontal and squamosal. 
Only the boundaries of the quadrato-jugal are generalized. The premax- 
illary is a little crushed to the left, and the most striking feature is the 
low-set position of the squamosal, which is but little if at all exaggerated. 
Cf. figure 6. 

Not only is the present type one of the most complete of 
fossil turtles, but more than any other known specimen of 
jRrotostega, it permits exact comparison with Archelon, being 
for the most part free from the crushing which so often 
obscures the characters of the otherwise line material from the 



G. P. Wieland — On Marine Turtles. 105 

Kansas chalk. Owing to this freedom from crushing, it has 
been possible to restore with approximate accuracy the outline 
of the skull, carapace, and plastron, although all the elements 
of the entire skeleton were dissociated during erosion from 
their matrix, those recovered being as follows : 

Skull, with lower jaw, — nearly complete, one squamosal and 
certain minor portions only missing. 

Carapace: Nuchal; first .to fourth neuralia ; pygal ; fairly 
complete series of ribs ; first and second marginalia of both 
sides, with third and fourth of right side. 

Plastron : Alae of the entoplastron ; hyoplastron of right 
side ; hypoplastic and xiphiplastra complete. 

Shoulder girdle : Both humeri and the procoraco-scapulars, 
with coracoid of right side only. 

Pelvis : Only the right ischium missing. 

The chief parts lacking, therefore, are the radius and ulna, 
the femora, and the bones of the hands and feet. 

The dissociated elements of the cranium, as brought together 
and mounted with the lower jaw, afford the most satisfactory 
representation of a Protostegan skull thus far seen. In fact, 
the result displayed by photographic figure 1 must be of 
nearly the true form, since in addition to the presence of the 
lower jaw and nearly all the exterior elements, the main 
outline is further confirmed by the practically complete pala- 
tines, pterygoids, and quadrates. Only in the interior of the 
skull are clear characters lacking; for instance, the descending 
process of the parietals, noted by Dr. Hay in Protostega 
advena, cannot be observed. 

In the main, the present fine cranium merely serves to 
corroborate the characters of the Protostegan skull, as already 
determined, and to bring out more clearly the major differences 
from Archelon. Thus, the strongly decurved beak and the 
upturned nares of the latter genus are absent, the outlines 
being more like those of other sea-turtles, with the orbits fairly 
well forward. The low-set squamosal, which certainly sent up 
a process along the posterior border to meet the parietal, how- 
ever, is a family characteristic. The general outline reminds 
one not a little of the skull of Colpochelys Kempii Garman. 
Interiorly, there is no great conelike palatal projection of the 
vomer, as seen in Archelon. 

While the present species is here defined as new, there is no 
very marked character not possessed by Protostega gig as, spe- 
cific differences being mainly exhibited by the smaller limbs 
and the heavier carapace and plastron now to be described. 

Carajyace. — Hitherto it has not been possible to gain a 
satisfactory picture of the shell of any species of Protostega. 
All the specimens known have either lacked a large part of the 



106 



G. R. Wieland — On Marine Turtles. 



neuro-pleural series, or they have been so badly crushed as to 
render the general form and structure more or less doubtful, 
as in the case of the Carnegie Museum specimen that yielded 
such clear testimony to all the characters of limb organization; 
and even in the present instance the evidence is not so convinc- 



Fig. 2. 




Figure 2. — Protostega Copei, 1/7 natural size. Carapace of the type as 
mounted in the Yale University Museum, from little crushed but dissociated 
elements lacking those portions marked by an ( x ) or else given in dotted 
outline. The portions actually present thus include a fine nuchal (with a 
nether process), the four first neurals and pygal with the three first margin- 
als perfect, together with eight pairs of pleuralia. The disk is then correctly 
indicated ; but concerning doubts as to the existence of a large ninth pair 
of ribs in contact with the marginals consult the text. 

ing and complete as in Archelon, where the series of ribs is not 
only entire, but articulated. Nevertheless, owing to absence 



G. B. Wieland—On Marine Turtles. 107 

from crushing, splendid conservation of all surface features, and 
the presence of a nearly complete series of ribs, with the 
nuchal, the anterior neurals. and important marginals com- 
pleting all the frontal border, the present carapace must be 
regarded as a magnificent specimen. In fact, the only 
structural point in the restoration here given, which awaits 
confirmation or disproval by future discovery, is the degree 
of development exhibited by the tenth pair of ribs. They are 
represented free, as in Archelon in figure 2, and such ribs 
extending out to meet the final marginal are regarded as a prob- 
able family distinction. It may be, however, that placing 
the proximal portion of the right fifth pleural as assigned and 
then restoring a sixth and a seventh pleural on the leftside are 
not warranted. In such a contingency two suppositions are, 
therefore, open, as follows : (a) The pair of ribs shown as the 
tenth may be really the ninth, and the true tenth pair of ribs 
may not be present, perhaps being only slightly smaller than 
those shown in the restoration, but passing out to meet the 
marginals. Such being the case, the only error made is in 
placing the pleuralia, from the fifth pair of ribs on, one number 
too far back ; (h) An unrecovered tenth pair of ribs may be 
reduced, as in the Oheloniidse, and may not have passed out to 
the final marginal. In this case, the post-fifth pleuralia would 
not only be one number too far back, but the carapace would 
be as here represented several centimeters too long. If either 
error has been made, the former seems by far the more 
probable. 

The neuralia are heavier than in Protostega gigas, and 
form a strong unbroken mid-ridge of normal Testudinate form 
in sharp distinction to the epineural grooving and anomalous 
structure of Archelon. Though it is to be noted that on the 
first neural, the second and third, evidently on the fourth and 
fifth, and probably on the missing eighth and ninth, there is 
a strong accentuation of the mid-ridge, suggesting the appear- 
ance seen in Toxochelys Banri. In the latter form, however, 
this feature is due to discrete epineural ossicles, while in Pro- 
tostega Copei there is no evidence of osteodermal elements. 
Nor are there any hornshield groovings ; on the contrary, the 
evidence is always to the effect that the Protostegidse were 
enveloped in a leathery hide. 

Fortunately, the nuchal is sufficiently complete to show the 
entire outline as a heavy normal element much as in Osteo- 
pygis, except that a prominent nether process is present. The 
first marginals, which are rather short and flat elements, are 
quite complete, as is also the rather long second marginal 
of the left side, with the distal half of that of the right side. 
On the latter may be traced with precision, one after the other, 



108 



G. R. Wieland — On Marine Turtles. 



the articular digitations and grooves corresponding to those of 
the next member of the marginal series, namely, the third, 
which is also complete and is followed by a fourth in equally 
good preservation. I am thus explicit, because it is important 
to note that all the anterior border of the carapace is indicated 
with certainty, and shows the presence of a peculiar upturning 

Fig. 3. 




Figure 3. — Protostega Copei. Plastron of the type x 1/7 nearly. The 
hypo-xiphilastrae are especially well conserved, and the left hyoplastron in 
fair condition. The right hyoplastron and portions of the T-shaped epiplas- 
tron are not present ; and as indicated in the figure, the ends of most of 
the digitations of all the elements are missing. Nevertheless the plastral 
form is quite accurately indicated in its entirety by the original specimen 
as here outlined. 

of the portion of the carapacial edge formed by the junction of 
the third and fourth marginals. Evidence is here furnished 
that the strong anterior and much upturned prolongation of 
the third marginal of Archelon was an articulating portion 



G. JR. Wieland — On Marine Turtles. 



109 



rather than a spine ; and use has been made of this fact, 
although it might have remained in doubt but for the exact 
testimony of the present "fine specimen. The interior borders 
of these marginals, however, do not have the strong digitation 
seen in Archelo/u being only slightly serrated. 



Fig. 4. 




Figure 4. — Protostega Copei. Shoulder girdle with, humeri and the 
pelvis of the type x 1/7. The only portions restored are the left coracoid 
and the right ischium. These elements belong to the very same individual 
as the skuil, carapace and plastron shown in figures 1-3. [It is likely that 
consonant with the heavy carapace the limbs were shorter than in P. gigas, 
and that the coracoids did not actually come into contact with the ecto- 
pubis.] 

Protostega Copei was not so orbicular a form as ArcJielon 
ischyros, and its plastron was relatively much shorter. As to 
the elements other than those now described, it appears neces- 
sary only to state that the accompanying illustrations suffi- 



110 



G. B. Wieland — On Marine Turtles. 



ciently exhibit not merely their characters, but their chief 
dimensions. Certain measurements, however, are appended, as 
follows : 

Length of cranium from snout to condyle (accurate) __ 24'0 cm 

Greatest width of cranium across the condyles (close).. 20*0 

Length of lower jaw (on median line) 15*0 

Length of carapace (estimated) 80- + 

Greatest length of plastron 63*5 ± 

Greatest width of plastron . 66*0 

Greatest thickness of plastral elements measured through 
their centers of ossification : 

Entoplastron __ 15 ,mm 

Hyoplastron 23* 

Hypoplastron _. 20* 

Xiphiplastron 13* 

Fig. 5. 




Figure 5. — Protostega (Archelon) Marshii. Eight third marginal x|. Dorsal 
view to the left, and ventral view showing large pit for the second rib to the 
right. [Dactylations should be represented as distinctly acuminate, but not 
much longer. J Observe, that as proven in Protostega Copei, articulation with 
the second marginal was formed "by the long upward and forward projection, 
resulting in abrupt change in direction of the marginal line at the humeral 
notch, as revealed further by the figures of the entire skeleton. 

Protostega (Archelon) Marshii Wieland 11 (Figure 5). 

In this Journal for April, 1900, 1 gave a brief description of 
a new species of Archelon, which was based on the portion of 
a skeleton collected by me in 1898, on the left bank of the 
Cheyenne River. Until further material is found, as will with 
certainty transpire, the fragmentary condition of the present 
remains will scarcely justify much work upon them by a pre- 



G. R. Wieland — On Marine Turtles. Ill 

parator ; nevertheless, the species represents an important 
fossil type. 

A reexamination, of the specimen confirms the characters 
given, namely, a relatively short humerus and great thickness 
of the plastron, the latter being half as thick again as that of 
Archelon. As it is probable that the present turtle was not 
quite as large as A. ischyros, type (3*4 m.=ll feet long), 
however, it may be that its plastron was proportionally twice 
as thick as that of the latter species. 

A fine third marginal of the right side shown in figure 5 is 
also present, and with it are articulated the keels of the more 
fragmentary fourth and fifth marginals. These elements are 
of much the same form as in A. ischyros, type, and although 
relatively heavier than in that species do not show the great 
disparity in weight noted in the plastron. More obvious differ- 
ences of taxonomic bearing, however, are exhibited by a frag- 
mentary, though otherwise finely conserved neural from near 
the middle of the neural series. This lacks the groove so 
highly characteristic of Archelon, and has a strong and continu- 
ous median ridge precisely like that in Protostega Copei. It 
therefore becomes necessary to transfer the present species 
from the genus Archelon to Protostega, where it holds a posi- 
tion of importance, as exhibiting not only the continuation of 
the latter genus from the Niobrara into the Fort Pierre, with 
marked increase in size, but, as far as we know, represents the 
closest structural approach of the genus Protostega to Archelon, 
thus far observed. 

The Mounted type of Archelon ischyros (with Figures 6-12, and 

Plates II-IV). 

All the material thus far referred to the genus Archelon has 
been discovered and collected by myself during the past four- 
teen years. The original type of Archelon ischyros was found 
in the brakes of the south fork of the Cheyenne Kiver, about 
five miles west of the mouth of Rapid Creek, Custer county, 
South Dakota, in August, 1895. Though a remarkably com- 
plete fossil, it lacked the skull, which, however, was supplied 
by an excellent younger specimen with a fine cranium and the 
lower jaw in place, obtained in 1897. This is here shown by 
the photographic drawing, figure 6. In 1898, the related type 
Protostega {Archelon) Marshii was procured from the same 
horizon as the specimens of Archelon, but on the east bank of 
the Cheyenne, in the Pine Ridge Indian reservation. Later 
still, in 1902, a large individual of A. ischyros, nearly identical 
in size with the original type, was collected at a point several 
miles farther south, on the west bank of the Cheyenne. This 



112 G. R. Wieland — On Marine Turtles. 

specimen lias been of considerable value in showing the more 
important carpals in natural position, and in yielding additional 
linger bones and the epiplastron. A well-conserved lower jaw 
with fully coossitied rami accompanies it. Like the type, it 
was completely imbedded in one of the lenticular masses of 
marl or clayey limestone common in the Pierre, but as this 
was much checked by cleavage planes there has been consider- 
able shifting of parts. The specimen, while good, is not com- 
parable to the original type, which, barring the lack of the 
skull (destroyed by erosion), is one of the finest of all great 
fossil vertebrates. It was but little crushed, and nearly all 
the parts present were in their normal position. 

In addition to the examples mentioned, fragmentary portions 
of other specimens were obtained at different times, all per- 
taining to the uppermost one hundred feet of the Pierre, and 
all from within an area of about eight square miles. The best 
skeletal conservation was found in the bluish clays of the 
upper thirty feet of the Pierre, as covered by the Oligocene 
overlap in the Cheyenne Kiver valley ; but most unfortunately 
a broad Oligocene river, the clearly marked bed of which I 
definitely located west of the Cheyenne, scooped out of these 
Pierre strata exactly the portion that must once have con- 
tained the most numerous and the best turtles, as indicated 
by frequency of occurrence as well as fine conservation on both 
banks of the old Oligocene valley. 

The various specimens of Archelon have been made the sub- 
ject of five contributions to this Journal. 6) 8> "• ,2 ' 14 . These par- 
tial descriptions have been repeated in a summarized abstract, 
with certain additional interpretations and views, in Dr. Hay's 
great volume on the " Fossil Turtles of North America." 17 
It does not therefore seem necessary again to repeat the 
preliminary descriptions, except in so far as needed to call 
attention to inaccuracies disclosed by the final mounting of the 
type specimen, together with the great additions to our knowl- 
edge of Protostega made during the past half dozen years. 
It is the present purpose to give in concise form the features 
of this greatest of marine turtles disclosed by new discoveries 
and by mounting, which always sheds new light on the char- 
acters of a fossil vertebrate ; and especially to give the facts of 
classificatory value, together with a discussion of relationships. 

The Skull (figure 6). — It is to be hoped that a second skull 
may be recovered. Thus far only the type skull and one addi- 
tional lower jaw have been found ; hence, it is not possible to 
add to the earlier descriptions except in wholly minor points 
of interpretation. Moreover, abstracts of these descriptions 
with figures are given in admirable form in Hay's work previ- 
ously cited, a work which every student of the Testudinata must 



G. R. Wieland — On Marine Turtles. 



113 



find indispensable. It may, however, be stated that in the res- 
toration, Plate IV, the skull undoubtedly shows the exact 
proportions of the supra-occipital, this being an improve- 
ment upon the earlier figures. 

The Vertebral Column. — The fact that the vertebral column 
of Archelon is so nearly complete and uncrushed, with nearly 
all the elements in undisturbed natural position, gives to the 
restoration high value. Only the four proximal cervicals and 
a few of the smaller caudals from near the tip of the tail are 
missing, while from the eighth cervical to the fifth caudal, 

Fig. 6. 




Figure 6. — Archelon ischyros. Skull of cotype shown about 1/8 natural 
size. The bounding sutures of all the exterior elements may readily be 
distinguished. [The restored supra-occipital crest, perhaps, is shown. too 
slender.] The low set squamosal is, in comparison with P. Copei and P. gigas, 
seen to be a family character of the Protostegidee. 

inclusive, all the vertebrae are complete and in natural contact. 
Further, in the large specimen collected in 1902, the fourth 
cervical is present and of the normal or cyrtean form common 
to all marine Testudinata. 

In Archelon, the cervicals succeeding the fourth are coelo- 
cyrtean, with the inferior sides of the centra heavily double- 
keeled. The valley between the keels is broad and shallow, 
Neither in form nor size is there much variation in the distal 
cervical centra. Aside from strength and great size, the dorso- 



m 



G. R. Wieland — On Marine Turtles. 



sacral series presents no marked peculiarity, while the caudal 
series is rather short. The arches of the first eight to ten 
candals are free, but those of the- remainder of the series are 
strongly fused to their centra. It is not likely that the tail 
could have been so freely moved as might be implied from the 
over strong curve in the restoration. 



Fig. 7. 




Figure 7. — Archelon ischyros, x 1/36. Dorsal view of original type. 
It is not necessary to give in detail the exact outline of the restored portions. 
The skull is outlined from a cotype, and the right hind flipper shown in 
the normal outline was not present beyond the proximal half of the tibia 
and fibula, having, as explained in the text, been bitten away in the early 
life of the turtle. 

Note on the mid-line following the nuchal seven larger plates followed by 
four smaller ones and then the pygal. All these eleven plates intervening 
between the nuchal and the pygal are adjudged to be epineurals seated on 
the greatly reduced neural series which did not, as in all other turtles except 
Dermochelys, normally come into view at all. 

Observe the nine pairs of fully developed ribs, of which the first large pair 
is the second, the small first pair not coming into view. Infra- and supra- 
peripherals not indicated. 

The Carapace. — The nearly orbicular form of the carapace 
is a striking feature. The nuchal is very large and very thin, 



G. B. Wieland — On Marine Turtles. 115 

especially in the lateral portion, indeed so thin as to suggest 
the necessity for strengthening by overlapping dermal ossifica- 
tions. The "anterior edge is strongly concave, being sharp, not 
rounded, in the middle region. The nether or cervico-neural 
articular process is prominent ; it takes the form of a heavy 
trapezoidal ridge, longest in front, with a keel-like buttress 
radiating from all four corners. 

The medial elements of Archelon are anomalous and require 
discussion as well as description. At first sight one would 
certainly say that there is a series of seven larger neurals fol- 
lowing the broad thin nuchal, with four much smaller neurals 
preceding a single pygal element and making eleven neurals 
in all. On closer inspection, however, it is found that despite 
the fact that the neural region of Protostega is of simple and 
normal structure there is in Archelon a more complex arrange- 
ment of parts than in any turtle thus far discovered — a condi- 
tion moreover that has a more distinct bearing on the meaning 
of the Dermochelan anatomy than any other thus far observed 
in fossil turtles. 

In my original description I stated that " The medial plates 
unite very imperfectly by means of loosely doubled interlock- 
ing sutures and overlapping digitations grading into frequent 
free spines [spine-like projections] posteriorly. These digita- 
tions are mostly long, thin and ribbon-like, and produce a 
junction quite different from the usual suture. In many cases 
there is an appearance such as would result if the digitations 
of the one plate had lain upon the surface of the adjoining 
plate when it was in a plastic condition and thus raised rounded 
ridging about their edges. The order of the digitations and 
their size is rather regular." It was also explained that the 
carapace was very thin on the midline and that at a break expos- 
ing the section back of the sixth rib there were thin layerings. 
And it should now be added that, bearing in mind that in the 
mounted specimen the dorsal vertebrae remain articulated as 
originally, I am, perhaps, censurable for not having the cara- 
pace sawn through at the line say of the third, sixth and eighth 
dorsal centra. However this may be it was not done, and, 
awaiting further specimens, the type specimen in which all 
details are certainly present must yield as its only quota of 
new fact the superficial details. Indeed, were it not for the 
accidental fact that back of the ninth medial element two 
prepygal median elements are missing, and yet that there is 
continuity of tiie carapace, it would likely escape us that two 
layers of bone are present on the midline. At the point where 
these elements have become disarticulated one can see that the 
pleurals expand broadly beneath the median plates, but we 
cannot tell to what extent they replace or crowd the neurals, 



116 



G. M. Wieland — On Marine Turtles. 



which are evidently small posteriorly. Immediately back of. 
the nuchal, however, there does not appear to be an underly- 
ing neural, and one may say with certainty that all the anterior 
neurals at least are very greatly reduced. It is, too, equally 
clear that these mere thin sheets of underlying bones that 
can be referred to neurals, so far as they were seen in 
fracture sections, have had their function taken over by the 

Fig. 8. 




Figure 8. — Archelon ischyros, x 1/36. Ventral view of type. Compare 
legend of the preceding figure. Note that all the elements of the plastron 
here shown occupied their normal position as originally recovered. Only 
the doubtful position of the omitted epiplastra prevents final accuracy. 
Note also the great length, of the plastron ; it is nearly as long as the cara- 
pace. Observe the small size of the femoral notches. The rib pair passing 
beneath the hyo-hypoplastral suture is the fifth. 

outer dermal series, which is continuous from nuchal to pygal 
and thus corresponds to the neural keel of Dermochelys. 

On the nuchal itself no additional elements were observed, 
but, as just stated, one may suppose from the great reduction 
in thickness such may well have been present. Indeed, it is 
not impossible that the posterior end of the nuchal extended 
beneath the first of the median supra-neural elements, and 



G. R. Wieland — On Marine Turtles. 117 

that as a consequence our restoration is thus some ten centi- 
meters too long. 

Summarizing then : there are to be seen on the midline 
apparently overlapping all the neurals and the proximal bor- 
ders of the laminae of the nine pairs of normal ribs which 
pass out to meet the marginals, a series of eleven thin supra- 
neural elements simulating in size and outline a neural series. 

These supra-neurals form a distinct median keel and are of 
distinctly quadrangular outline all the way back to the eighth 
and ninth, which are short on the median line, but nearly as 
broad as the others. The digitate character of the sutures 
between the successive members, but more particularly of the 
pleural overlap, has already been commented upon. All the 
outlines are quite exactly shown in my original figure (refer- 
ence 6, Plate VI). A dominant feature of the midline is a nar- 
row median groove which extends from the second to the 
seventh member inclusive and is most pronounced in the sec- 
ond and fifth. In the mid-region of each neural enumerated the 
groove is somewhat widened and deepened, sending out a 
radially ranged series of nutrition furrows or striations which 
form a dominant sculpturing of the mid-region of the cara- 
pace. Some further horny or even ossified elements may have 
occupied the mid-region of these shields. The aspect of the 
neural keel is thus seen to be different from that of any other 
turtle. 

The supra-neurals of Archelon, be it noted, vary distinctly 
from those of Toxochelys in which the series is not contin- 
uous and corresponds to a normal series of vertebral horn- 
shields. 

In Arehelon, however, a leathery hide must have been pres- 
ent, with a system of keels of the usual number, as denoted 
by dermogene ossifications rather than hornshields ; these will 
be treated more fully later on. An ossicle like the supra- 
neurals of Toxochelys was found by Hay accompanying 
Protostega advena, but its derivation was left in doubt. 

The pleural investment of the ribs occupies only the proxi- 
mal fifth of their length. The free ends of the ribs are thus 
the dominating feature of the carapace. They are very heavy, 
in compensation for the light to almost paste-board thickness 
of the carapacial shield. The first rib is small and more or less 
curved and flattened. As in Protostega, it passed well to the 
front beyond the expansion of the first pleural, and may have 
supported either the thin posterior nuchal ala or possibly some 
osteodermal element. 

It is here necessary to note that the type specimen remained 

packed, partly in the matrix, from 1898 until 1906. Owing 

primarily, however, to a luckless defect in my field notes, 

Am. Jour. Sci.— Fourth Series, Vol. XXVII, No. 158.— February, 1909. 
9 



118 



G. i?. Wieland — On Marine Turtles. 



which rendered a lapse of judgment easy, it was thought for 
a time that the crushed coracoid of the specimen collected in' 
1902 was a heavy first rib. 

As a result of this misconception, together with the uncer- 
tainty regarding the carapace that had long existed in the 
mind of every student of the Testudinata, I published the 

Fig. 9. 




Figure 9. — Archelonischyros, x 1/36. Ventral view of the type witli plastron 
removed. Compare legends of the two preceding figures. Eecall that as 
originally collected the vertebral column from the fourth cervical on was 
found normally articulated and complete all the way to the smaller caudals, 
a few of which were missing. Observe that the coracoids pass all the way 
back to the ectopubes. Note in the hind paddle the large size of the tibiale 
and fibulare, and the pisiformoid development or functioning of the fifth 
carpal. 

erroneous figure of my paper of 1903. 14 But why Dr. Hay has 
reproduced this in his volume/ 7 I fail to understand, as I spe- 
cifically declared the figure to be a mistaken one several years 
ago. I can only regret that the original figure of A. ischyros 6 
was not used, and still more deplore the fact that the labor of 
preparation on the type was not sufficiently advanced to permit 



G. R. Wieland — On Marine Turtles. 119 

the offer to Dr. Hay of a photograph of the mounted skeleton 
in time for use in his volume. 

The tenth ribs verify the original and excellent figure of the 
carapace in an important detail. In that figure, these ribs are 
both shown as distally restored to a length indicating support 
of the last marginal, were that element present. This restora- 
tion is correct, the right tenth rib having since been found 
complete, so that the length of the entire series is now abso- 
lutely known ; and it should here be emphasized that while 
some of the ribs had disintegrated on one side or the other of 
the carapace, there is not a pair in the succession, from the 
first to the tenth inclusive, that lacks either a right or a left 
member complete to the tip. 

This functional development of the tenth rib is unique in 
the Thecophora. It denotes either a more primitive condition 
or a restrengthening of this element in compensation for a 
carapacial shield not only in process of reduction, but probably 
also of replacement by an external dermal series corresponding 
to the usual Testudinate keels or lines of longitudinal develop- 
ment. 

The Marginals. — The marginal series of A. ischyros, type, 
is considerably restored in the figures given (Plates II-IV), but 
not hypothetically so. There are present in sufficiently good 
condition for the determination of all the main features, margin- 
als referred to ribs extending all the way from the second to 
the eighth or ninth rib. Further, the fine third marginal is 
present (cf. figure 5), which in Protostega (Archelon) Marshii 
is suturally united with the fourth to sixth, the latter species 
having marginals of quite the same form as in A. ischyros, type. 
In the additional specimen of A. ischyros obtained in 1902, the 
seventh (?) marginal is also present, while the first to fourth 
are positively known in Protostega Copei (cf. figure 2). 
Hence, remembering the functional tenth rib, it may be defi- 
nitely stated that each rib beginning with the second bore a 
marginal and that the pygal marginal, the only member of the 
peripheral series not recovered in any of the larger forms 
of the Protostegidse (it is present in P. advena), was thin, 
short, and broad, and must have had the form shown in the 
restoration (Plate II). 

The noteworthy and strange feature of the marginals is the 
strong digitation of both the superior and inferior plates and 
also of the anterior elbow formed by the outer border of the 
third marginal. Were it not for the fact that in P. Copei the 
junction with the second marginal is definitely shown, it would 
scarcely be suspected that the true articulation of the third 
marginal with the second in A. ischyros, type, takes place by 
means of its long spinelike extension, which projects upward 



120 



G. R. Wieland — On Marine Turtles. 



and forward ; for even in that portion of the outer border next 
to the humeral notch there are blunt spines. 

Did the spines of the marginals articulate with one or more 
carapacial and plastral rows of dermal ossifications, and thus 
afford the nearest approximation to the osteodermal mosaic of 
Dermochelys yet discovered by adding dermal ossification on all 
the keel lines, that is the neural keel, the pleural keel, the supra- 
marginal, and marginal 




keel! 



I 



Figure 10. — Archelonischyros x^. 



A large 
dermal ossicle found in connection with the 
plastron of the original type, but possibly 
dorsal. O, the outer view ; S, sectional view 
showing the extreme thinness of the element. 
Note that the asymmetry of this element and 
its dactylate border indicate not only connec- 
tion with other dermal elements, but the 
probability of the presence of entire series of 
such elements. 



believe such to 
be the case, for at least 
tw T o elements referable 
to a similar additional 
series corresponding to 
the supra-pleural keel of 
Dermochelys have been 
recovered. Nor is it 
strange that more have 
not been obtained. A 
fine shark's tooth per- 
taining to a scavenger 
species related to Lam- 
na was found with the 
type, and clearly indi- 



cates that all dermal por- 
tions loosely affixed to 
other elements of the 
carapace or plastron 
must have been pecu- 
liarly liable to disassocia- 
tion. In what other than a supra-marginal or infra-marginal 
position is it possible to place the thin and distinctly asym- 
metrical element shown in figure 10 ? It represents an 
integral part either of the carapace or of the plastron. Further, 
the likewise unique element shown in connection with the 
marginal in figure 11 can not be interpreted as in other 
than a natural position. From the fact that it is digitate all 
around and slightly asymmetrical, it may be inferred that a 
series of such elements lay inside of, and articulated with, the 
superior borders of the marginals, alternating quite regularly 
with them in about a double number, and that beyond this 
space a second much thinner supra-marginal series was present. 
The space between the latter and the midline of the cara- 
pace, where, as has been already seen, the presence of a median 
row of supra-neural or in part osteodermal elements is demon- 
strated, may or may not have been continuously occupied by 
ossifications. In any event, there are the seven dorsal keels, 
as in Dermochelys. On the plastral side, direct evidence of 



G. JR. Wieland — On Marine Turtles. 



121 



dermal elements is lacking, although, an agreement with the 
five plastral keels of the leatherback may be conjectured. 

The Plastron (figure 8). — The nearly perfect plastron of the 
type has suffered somewhat during collection, by its removal 
from an exceedingly hard marl matrix into which the numer- 
ous and often interlocking spines of the mid-plastral region 
penetrated. Nevertheless, as finally mounted, the plastron 
may be said to be in splendid condition. All the central 

Fig. 11. 




Figure 11. — Archelon ischyros. Left eighth or ninth marginal as found 
in conjunction with an additional element in a supra-marginal position. 
Shown % natural size. If the anomalous element is not a supra-marginal 
it must be referred to the supra-neural series just anterior to the pygal. A 
.slight asymmetry does not prevent, although it makes such a position less 
probable. It is easier to consider this element as having been found in a 
natural position, and as perforce thus accounting for the supra-marginal 
keel of the carapace of Dermochelys. 

portions of the large plates are present, with most of the spines, 
so that neither their size, form, length, or number, is ever in 
doubt. More important still, all the elements are but little 
crushed and, save the epiplastra, are present in their normally 
articulated position, just as they were figured in 1898. 9 In 



122 G. B. Wieland—On Marine Turtles. 

commenting on this figure, Dr. Hay states that a length of 
2100 millimeters is thus indicated for the plastron, making it 
larger than the carapace, which he considers impossible. 
Neglecting my measurement of the plastron, which was given 
as 2000+ millimeters, as well as the fact that the entoplastron 
is very plainly shown a little anterior to its true position, he 
also fails to note that the exact length of the carapace with, 
the nuchal in position had not been determined by anyone. 
His first premise is therefore unfounded and his conclusion a 
pure assumption. 

The important point, however, is that in the restoration, 
where the length of both carapace and plastron is definitely 
determined, the two are found to be nearly equal. In the 
dorsal view, the plastron appears a little shorter than the cara- 
pace, while in the ventral view, the enormous expanse of the 
plastron, greater by far than in Protostega Gopei and greater 
than in any other sea-turtle, entirely cuts out the carapace. 
Archelon ischyros was certainly a very singular marine form ; 
with its enormous size, huge plastron, and small femoral notch 
set far back, it had need of the great humerus, which by reason 
of form and musculature represents a powerful sea-type. 

The unique T-shaped entoplastron of the Protostegidse has 
had an interesting history. First called a nuchal by Hay in a 
Kansas specimen, it was left for Wieland 8 to determine con- 
clusively and figure both these elements in A. iehyros, although 
there was uncertainty whether epiplastra were present at all. 
For reasons that now appear trivial, being merely an imper- 
fectly indicated condition of overlap seen in the field, the excel- 
lent point of view developed in the paper just mentioned was 
abandoned for a time. 14 

Meanwhile the specimen of 1902 w T as obtained and was 
found to include still another puzzling bone, — the element 
that must be regarded as an anomalous epiplastron ; and still 
later the fine type of Protostega Gopei, here described, was 
discovered by Sternberg and acquired for the Yale collections. 
Thus was I enabled to determine finally that the nuchal and 
entoplastron noted in the paper of 1898- 8 were truly such. 
This correction appeared in the Annals of the Carnegie 
Museum of Pittsburg for 1906. 16 

That scarcely one of the naturalists interested in the Pro- 
tostegidse escaped from wrong conclusions as to the nuchal and 
entoplastron, is after all not surprising. Both elements are of 
a form not before observed, this being especially true of the 
entoplastron, which except in P. potens Hay shows no indica- 
tion of any ordinary type of epiplastral superposition or 
junction. 



G. R. Wieland — On Marine Turtles. 



123 



The epiplastra are doubtless of the form shown in figure 12. 
Referring to my first description, however, Dr. Hay thinks 
that the element figured must be the right, not the left 
member ; superposition would therefore not be of the out- 
turned Trionychoid type that 1 have supposed. Dr. Hay saw 
this element soon after it was collected, and is consequently in 
a position to judge ; nevertheless I think he errs and that the 
explanation of his opposite opinion is the condition he has 
observed in the entoplastron of P.potens. Moreover, I am 
not sure that he has correctly determined the hyo- and hypo- 
plasia in that turtle, for the elements he figures as xiphiplastra 

Fig. 12. 




Figure 12. — Archelon ischyros. Left epiplastron, x £. Ectal view on 
the right below and ental view on the left. On the right above, the ante- 
rior, and on the left, the posterior edge views of the recovered portion are 
shown. (There is no doubt that the restoration of the thin dactylate end is 
fairly accurate both as to form and size.) This element was not present in 
the original type, having only been observed once in all the history of the 
Protostegidas. 

I should certainly have called hyoplastra. In either case, 
however, P. potens, the type of which Dr. Hay was kind 
enough to show us, is a -quite different turtle from any of the 
foregoing, and the evidence it affords as to the form of the 
epiplastra is only negative and quite uncertain. It seems 
much better to accept the positive evidence at hand, which is 
to the eifect that if the element figured is the true epiplastron, 
it projected beyond the anterior border of the entoplastron and 
was borne on it quite as in Trionychids. But rather than risk 
finality in error, it has not been given a place in the restored 
type of A. ischyros. 



12i G. P. Wieland — On Marine Turtles. 

The hyo- and hyjtoplastra exhibit no very unusual features, 
except a great number of peripheral spines. The curved or 
somewhat boomerang-shaped xiphiplastra are of course more 
primitive than are the long and straight forms common to the 
Cheloniidse. 

That the plastral fontanelles appear to be of less area than 
is shown in figures of Protostega, is due more to the fact that 
the plastron under consideration is the best and most complete 
example known in the Protostegidse than to any marked vari- 
ance in proportions. The plastral resemblance in Protostega 
and Archelon is very striking, in view of other differences 
separating these genera. 

The Shoulder Girdle and Manus. — The marked feature of 
the huge shoulder girdle is the projection of the coracoid all 
the way back to the pubis, a feature also present in Protostega 
and common to the existing Eretmochelys. The most charac- 
teristic element in the shoulder girdle of Archelon is the 
humerus because of its distinctly thalassic type. 

The testimony as to the organization of the manus is reason- 
ably complete and aside from minor differences exhibits general 
agreement with that of Protostega. The centrale in the latter 
is, for instance, more distinctly angled. While all the carpal 
elements of either a right or a left flipper are present, only the 
principal bones of the carpus have been found in position or 
approximately so. It is only in the left flipper that bones 
from another specimen have been introduced, namely, carpale 
I, the intermedium, and the pisiform, which fortunately were 
found together in this supplementary specimen. The only 
element in doubt was the centrale, but this seems to have been 
of a rounder form than in Protostega. 

Of the metacarpals and phalanges, the majority are present 
and the proportions of the fingers are essentially those adopted 
in the restoration, although when a specimen is once found with 
these elements in place, as in the case of the Pittsburg Museum 
specimen, some slight modification of the present restoration 
may prove necessary. 

The important anatomical features of the front flipper then 
are: (a) Agreement with Protostega y. (b) general agreement 
with the Cheloniidae, the centrale exhibiting strong contact 
with metacarpal I, instead of exclusion from contact with this 
element by junction of the intermedium and carpale II ; 
(c) the comparatively slight modification and elongation of the 
phalanges for pelagic life, as contrasted with the much 
modified thalassic humerus. Although the latter is thus mod- 
ified, it lacks much of the strength exhibited by the paratha- 
lassic Dermochelan humerus ; for while the radial crest has 
shifted toward the middle region of the shaft, it has failed to 



G. JR. Wieland — On Marine Turtles. 125 

retain a strong pedestal affording a powerful and firm type of 
muscular insertion. Curiously enough, the earlier Niobrara 
Protostega was better provided in this respect, since its radial 
crest forms a distinct ala nearly as prominent as that seen in 
Dermochelys. 

This failure of Archelon to develop or retain, as the case may 
be, a prominent crest with stronger type of radial musculature 
may indeed indicate a certain failure to progress in swimming 
power and in resultant ability to follow the southward retreat 
of the great central Pierre sea. In fact, it was at just about 
this period of culmination in size of the Protostegidse that the 
Dermochelan line more successfully accomplished such a 
change, as shown by the Eocene Psephophorus, a turtle nearly 
approaching Archelon in size and having a strongly pronounced 
and very low-set radial crest. It is on such grounds, as much 
as by the possible destruction of the eggs of the young by 
marine or even by newly evolved mammalian enemies, that suffi- 
cient cause is surmised for the extinction of these most gigantic 
of all marine Testudinates. 

The Pelvic Girdle and Pes (figure 9). — The very perfect 
and uncrushed pelvis of the type was accompanied by the left 
femur, tibia, fibula, tarsals, and nearly all the metatarsals. On 
the right side, the femur is also present, with the proximal two- 
thirds of both tibia and fibula, which end in obliquely bitten 
off but healed surfaces. Both the femur and those mutilated 
elements are lighter and several centimeters shorter than the 
corresponding bones of the left side. In short, the evidence 
is conclusive and unmistakable that this animal had its right 
flipper bitten off when still young, and that as a result of this 
injury the remaining portion of the flipper was more or less 
arrested in growth by disuse. Such accidents are now and,' then 
noted in fossils. The type of Dromocyon vorax shows a broken 
lower jaw, subsequently reknitted, which was doubtless received 
in some raid on the young of Palmosyops, while a large per- 
centage of existing marine turtles have had their flippers more 
or less mutilated by predaceous fishes and sharks. 

I need not remind those familiar with the Testudinate osteol- 
ogy that the tarsal region of the sea-turtles is decidedly more 
variable in its organization than is the carpal region. Owing 
to this cause and to the failure to identify the excellently con- 
served tarsals with those of the crushed elements of Protostega 
gigas, it has not proved possible to orient the tarsals except in 
the most provisional manner. They are all free and heavy 
bones, and there is little doubt that all were present on the left 
side, however difficult and uncertain exact orientation may be. 

The metatarsals are more readily recognizable, the fifth be- 
ing much flattened and highly characteristic. Its distal half 



126 G. B. Wieland—On Marine Turtles. 

is largest, not smallest as in Protostega gigas. In closing this 
brief description of the flippers of A. ischyros, type, it should 
be emphasized that while there is marked resemblance to 
Protostega, it is only the resemblance of members of the same 
family, and that the chief variation is in the humerus and the 
pes. [The other region of marked variation is on the neural 
line ; the crania do not differ greatly.] 

The more important measurements of Archelon ischyros, 
type, are as follows : — 

[Weight of humerus, exactly 75 pounds = 34 kilograms.] 
Length of cranium from beak to occipital condyle 'd0 ,cm 

" four distal oervicals (estimated).. 35- 

" four proximal cervicals (present) 33' 

" ten dorsal centra measured on the ven- 
tral face 125* 

" two sacral vertebrae 9*5 

" eighteen caudal centra (estimated ; only 
a lew of the more distal members are 
absent) _ 67* ± 



Total absolute length from beak to tip of tail, 
from measurements on ventral face of centra 329*5 

(The corresponding total exterior measurement is not so 
accurately obtained, being slightly affected by pressure, but 
must have been 3*4 m = 11 feet.) 

Length of first dorsal centrum _•_ 8* cra 

" second " " 15 

" third " " . 16-5 

" fourth " " 16-5 

" fifth " " ., ]65 

" sixth " " 14*5 

" seventh" " 14' 

" eighth " ■" 10* 

" ninth " " 7* 

" tenth " " 6- 

Total length of ten dorsal centra 124*0 

Length of first sacral centrum 5 • 

" second " " 4'5 

" first caudal " 4*5 

" second " " ._ ._ 4*5 

" third " " 4-5 

fourth " " 4-0 

fifth " " 4-0 

sixth " " .. 3-0 



G. R, Wieland — On Marine Turtles. 127 

Carapace : — 

Greatest length on median line over slight curv- 
ature, as mounted. . 2 # 00 m 

Greatest absolute length 1-93 

Greatest width over partial curvature, as mounted. 2*18 
Width over curvature of second pair of ribs 2*10 

Plastron : — 

Absolute length on median line l*87 m 

Width on line drawn across the humeral notches 1*83 
Greatest width of hyo-hypoplastral fontanelle 

(that at hyo-hypoplastral suture) _. *87 

Distance between numeral and femoral notches 1*15 

Comparative measurements of the right and left femora, tibiae, 
and fibulae, showing the check on growth due to the loss of 
the right foot. 

Femora : — 

Eight Left 

Extreme length 43*5 cm 49' cm 

" distal width.. . 200 22*3 

Least antero-posterior thickness of shaft 7*2 »9' 

Tibiae:* — 

Greatest proximal thickness 13*3 15*0 

Fibuloe : — 

Eight Left 
Greatest proximate thickness 7 '2 cm 8 -3 cm 

Flippers : — 

Distance between glenoid cavities ._ 58* ± cm 

Extreme length of front flipper outstretched 

in a straight line (measure from glenoid 

cavity), _ . - 200* ± 

Extreme length from tip to tip of fully 

outstretched front flippers 458* ± 

Distance between acetabular foramina 28" 

Extreme length of fully outstretched hind 

flipper _ . 1 38- ± 

Extreme length from tip to tip of fully 

outstretched hind flippers 304- ± 

A few comparative measurements of Protostega and Arch- 
elon, clearly showing the differences in proportion to be 
expected in different genera, are the following : — 

*The distal ends of right tibia and fibula are cut away by a clean shearing 
bite : hence the measurements of the left side are the normal ones. 



128 G. E. Wieland—On Marine Turtles. 

A. ischyros (type) P. gigas (cotype)* 

Extreme length of coracoid 75 ,cm . _ . . 4o- cni 

Procoraco-scapular length 66* 30' 

Median line to tip of second 

rib 100- 

Extreme length of humerus 60- 

Length of ten dorsal centra 127' 

Extreme length of femur 49- 



a 



■<->■ 



tibia 33 

fibula 3] 

radius 30 

ulna 35 



56 

34 
.68 

27 
.20 
.20 

20 
.17 



Concluding Remarks. 

Not only will future field work reveal new members of the 
Protostegidse of the greatest interest, but quite all the skeletal 
features now in doubt must certainly be clearly observed as one 
specimen after another is collected. Indeed it can be freely 
predicted that but a very few years will be required to accumu- 
late the material demanding a second revision. 

Meantime it must be left to such further discovery to deter- 
mine, among various other features, what the exact condition 
of the neural line of JProtostega gigas really is, and whether 
this species and P. Copei do not really belong to separate 
genera ; for there is a distinct suspicion that the species of 
Protostegidse already known may really include a third genus. 
Evidently the marked difference in the structure of the cara- 
pacial midline between Archelon and Protostega Copei indi- 
cates a condition promising variations of the most striking and 
interesting character, to say nothing of the possibility of variety 
in the dermogene elements on the lines of the keels. These 
should very clearly be named in both Dermochelys and 
other turtles, the neural, pleural, supra-marginal, and marginal 
keels above, and the infra-marginal, hyo-hypoplastral, and the 
nether median or epi-xiphiplastral keels below. 

It is not presumable that there is any doubt as to the pres- 
ence of broad generic distinctions between Protostega gigas and 
Archelon, although the midline of these two forms may prove 
to be much more nearly similar than we now suppose. It is, 
however, a very striking fact that Protostega Copei and P. 
{Archelon) Marshii are both so much more nearly normal in 
their carapacial structure than is Archelon. This is to say, 
normal when we have in mind the great majority of turtles 
with normal neuralia such as the early Protostegids are shown 
to have. One might indeed suspect it possible from the strong 

*This is the splendid specimen, No. 1421, of the Carnegie Museum, 
Pittsburg. 



G. R. Wieland—On Marine Turtles. 129 

functional value of the epineurals.of Archelon that there are 
turtles in which following elimination of a true neural series, 
an overlying dermogene series like that of Archelon has 
dropped down into the neural position once more. 

The possibility of such cycles is, however, only hinted at. 
Taking the evidence at its face value, the important point is 
that Archelon, without having lost the power to develop an 
ossicular series, or perhaps in spite of having retained such a 
series, once had a closed carapace and plastron like that of 
modern turtles. Moreover, the earlier Niobrara Protostegas 
include the primitive forms like P. Copei and doubtless P. 
advena with well-developed neurals, and with far less of osteo- 
dermal development than in the later Archelon. 

It is thus seen that of the two camps which have attacked 
the difficult and highly attractive problem of the origin of 
Dermochelys, those favoring the view of a close relationship to 
other turtles and a comparatively recent origin have rather the 
best of the argument. We have had oh the one side Cope, the 
earlier Dollo, and Hay advocating an ancient and remote origin 
of Dermochelys / while on the other, Baur, the later Dollo, 
and Wieland have believed Dermochelys a highly specialized 
descendant of true turtles, and hence of modern or relatively 
recent derivation. That the latter of these hypotheses more 
nearly expresses the final truth is now evident ; though both 
contain elements of truth, and are by no means so remote as 
they at first sight appeared to be. 

Thus, just as Dollo deserted the one camp for the other, so 
Hay has gradually developed, not to say modified, his premises 
to a point where they adjoin our own. "While not absolutely 
closed, therefore, and still lacking the testimony of many forms 
yet sure to be discovered, this famous controversy of the biolo- 
gist as to the origin of the " leatherback " is now nearly elimi- 
nated. Nor is it too much to say that it has proved quite as 
fruitful throughout as the broader but scarcely more profitable 
question of the origin of the Testudinata. 

References. 

1 — 1872. Cope, E. D. — A description of the Genus Protostega. Proc. Am. 

Phil. Soc, Phila., vol. xii, p. 422. 
2. — 1875. Cope, E. D. — The Vertebrata of the Cretaceous Formations of 

the West. Kept. U. S. Geol. Surv. Terr., vol. ii, pp. 99-113, pis. 9-13. 

Washington. 
3. — 1884. Capellini, G. — II Chelonio Veronese Protosphargis Veronensis. 

R. Accad. dei Lincei. 36 pp., 7 pis. Rome. 
4. — 1889. Baur, G. — Die systematische Stellung von Dermochelys Blain- 

ville. Biolog. Centralbl., Bd. ix, p. 189. 
5. — 1895. Hay, O. P. — On certain portions of the skeleton of Protostega 

gigas. Field Columbian Mus., Pub. No. 7 (Zool. Ser., vol. i, No. 2), 

pp. 57-62, pis. iv and v. 



130 G. R. Wieland— On Marine Turtles. 

6.: — 1896. Wieland, G. E. — Archelon ischyros : a new gigantic Cryptodire 

Testudinate from the Fort Pierre Cretaceous of South Dakota. This 

Journal, vol. ii, p. 399, pi. vi. 
7.— 1897. Case, E. C. — On the osteology and relationship of Protostega. 

Jour. Morph,, vol. xiv, pp. 21-60, pis. 4-6. 
8.— 1898. Wieland, G. E.~ The Protostegan Plastron. This Journal, vol. 

v, p. 15, pi. ii. 
9. — 1898. Capellini, G. — Le Piastre Marginale della Protosphargis Veron- 

ensis. E. Accad. de Scienze dell. '1st. di Bologna. 
10. — 1898. Hay, O. P. — On Protostega, the systematic position of Dermo- 

chelys, and the morphogeny of the Chelonian carapace and plastron. 

Am. Nat., vol. xxxiii, pp. 929-948. 
11.— 1900. Wieland, G. E.— The Skull, Pelvis, and probable relationships 

of the Huge Turtles of the Genus Archelon from the Fort Pierre Cre- 
taceous. This Journal, vol. ix, p. 237, pi. ii. 
12. — 1902. Wieland, G. E. — Notes on the Cretaceous Turtles, Toxochelys 

and Archelon, with a Classification of the Marine Testudinata. This 

Journal, vol. xiv, p. 95. 
13. — 1902. Hay, O. P. — Bibliography and Catalogue of the Fossil Vertebrata 

of North America. Bull. U. S. Geo. Surv., No. 179, p. 439. 
14.— 1903. Wieland, G. E.— Notes on the Marine Turtle Archelon. I. On 

the Structure of the Carapace. II. Associated Fossils. This Journal, 

vol. xv, p. 211. 
15. — 1906. Wieland, G. E. — Plastron of the Protosteginae. Ann. Carnegie 

Mus., vol. iv, pp. 8-14, pis. i. ii. 
16. — 1906. Wieland, G. E. — The Osteology of Protostega. Mem. Carnegie 

Mus., vol. ii, pp. 279-298, pis. xxxi-xxxiii. 
17.— 1908. Hay, O. P.— The Fossil Turtles of North America. Carnegie 

Inst. Washington, Pub. No. 75, pp. 189-208. 



Washington — Submarine Eruptions of 1831 and 1891. 131 



Art. YIII. — The Submarine Eruptions of 1831 and 1891 
near Pantelleria / by Henry S. Washington. 

Introduction. — Of the regions which are noteworthy for 
submarine eruptions, that part of the Mediterranean between 
Sicily and Tunis has become classic. In this broad, deep strait 
lie the two wholly volcanic islands of Pantelleria and Linosa, 
which undoubtedly began with submarine eruptions, but whose 
volcanic activity seems now to be quite extinct, or almost so. 
That the vulcanicity of the district is not yet at an end has been 
manifested several times during the nineteenth century, in the 
years 1831, 1815, 1846, 1863, and 1891, as well as possibly in 
1801 and 1832. Of these the phenomena of the eruptions of 
1831 and 1891 were noted by competent observers, and material 
ejected by them has been preserved and studied. Specimens 
of the lava from both of these came into my possession and 
have been examined with the miscroscope and analyzed, in 
connection with a study of the rocks of Pantelleria which is 
being carried out for the Carnegie Institution. 

A map of the strait between Sicily and Tunis is given in 
fig. 1, based on the British Admiralty Chart 'No. 2158A, with 
a few additional data. The volcanic islands of Pantelleria, 
Linosa, and Graham are in solid black, and the site of Foerst- 
ner Volcano is represented by a straight line. The locations 
of the submarine eruptions of 1845 and 1846 are shown by 
small crosses. The site of the eruption of 1863, though 
reported to have been at that of Graham Island, is uncertain, 
and a note on the B. A. Chart No. 2127 indicates that it was 
searched for at the Hecate Patch. The hundred fathom line 
is marked by a dotted line, and two fifty fathom lines, inclos- 
ing the Adventure Bank and a small bank to the west, are 
marked by dots and dashes. Small " patches, " where the 
depth is less than 20 fathoms and often less than 10, are shown 
by small dotted areas. Outside the hundred-fathom line the 
soundings on the chart are seldom more than 400 fathoms, 
though two of 717 and 890 are marked between Pantelleria 
and Malta. 

It will be seen that the volcanic eruptions in general have 
originated in comparatively deep water and outside of the 
shelves which fringe the coasts of Sicily and Tunis, on which lie 
respectively the limestone islands of Malta, Gozo, Lampedusa, 
and Lampione, remnants of the early bridge between Italy and 
Africa, which were separated by faulting probably in Pliocene 
time. The known volcanoes, therefore, occupy the site of this 
fault, and the occurrence of the small, shallow " patches " in 
the continuation of the deep water to the northwest is interest- 



132 Washington — Submarine Eruptions of 1831 and 1891- 

ing in this connection. These areas of very shoal water are 
all small, and usually set in groups on a rather shallow "bank." 
In some cases, as the Keith reef, the bottom is so near the sur- 
face of the water that the sea breaks. Although many of 
these are covered with coral, the characters of their submarine 
topography (shown in detail for some on Chart No. 2127), 
which resemble those of shoals known to have been the sites 
of submarine eruptions, lead to the conclusion that these shal- 



FlG. l 




low patches have also originated in the same way, and that 
submarine eruptions have been of comparative frequency in 
this portion of the Mediterranean. That we have direct 
knowledge of so few of them need not cause surprise, when it 
is remembered that many of them probably gave rise to 
no solid or lasting island, as was true of Foerstner Volcano, 
and that their periods of activity were often short, so that, as 
was the case with the eruptions of 1845 and 1846, it is only by 
the rare chance of some passing vessel (generally the small 



Washington — Submarine Eruptions of 1831 and 1891. 133 

boat of some sponger or fisherman) that they are heard of. 
As Mercalli points out, also, onr knowledge of the eruptions 
of Vesuvius and Etna prior to modern times, situated on land 
and in populous districts, is so fragmentary and incomplete 
that our ignorance of the many possible submarine eruptions 
is to be expected. 

Graham* Island, 1831. 
Bibliography. 

The number of papers dealing with this eruption is fairly 
long. Johnston -Lavisf gives 28 titles and several others can 
be added to the list. The great majority date from the year 
of the eruption and those immediately succeeding, only one or 
two belonging to the latter half of the last century. The more 
important papers are those by H. and J. Davy, J H. Foerstner,§ 
C. Gemmellaro,|j F. Hoffmann, ^\ C. Prevost** and H.Abich;ff 
while general accounts are to be found in standard works deal- 
ing with vulcanology, as those of Landgrebe,^ Fuchs,§§ Mer- 
calli, || || Lyell,TfT and Bonney.*** 

The Eruption. 

The site of the eruption of 1831 was in lat. 37° V 30" IS". 
and long. 12° 42' 15" E., about 30 miles southwest of Sciacca, 
on the coast of Sicily, and 33 miles northeast of Pantelleria. 

The first signs of activity were shocks felt on board a vessel 
sailing over the spot on June 28, earthquake shocks being also 
felt in Sicily about the same time. During the first few days 
of July a fetid odor was perceived at Sciacca, and fishermen 
reported that the sea at the JNerita Bank appeared to be boil- 
ing and was covered with floating matter and dead fish. On 

* This small and ephemeral island has received seven names : Corrao, Fer- 
dinandea, Giulia (Julie), Graham, Hotham, Nerita, and Sciacca. That 
adopted here seems to be the best founded, as it was that bestowed by the 
first person who landed on it, Capt Senhouse, and is that used by English 
and many American writers. The Italians use either Giulia or Ferdinandea, 
and the Germans mostly the latter. 

fH. J. Johnston -Lavis, The South Italian Volcanoes, Naples, 1891, p. 105. 

X H. and J. Daw, various papers in Phil. Trans, for 1832 and 1833. 

§H. Foerstner, Tsch. Min. Petr. Mitth., vol. v, p. 388, 1883. 

|| C. Gemmellaro, Atti Ace. Gioen., vol. viii, p. 271, 1834. 

If F. Hoffmann, Pogg. Ann., vol. xxiv, p. 65, 1832. 

**C. Prevost, Mem. Soc. Geol. France, vol. ii, p. 91, 1832. 

f \ H. Abich, Vulkanische Erscheinungen etc., Braunschweig, p. 72, 1841. 

\%Gc. Landgrebe, Naturgeschichte der Vulkane, Gotha, p. 50, 1855. 

£§K. Fuchs. Vulkane und Erdbeben, Liepzig, p. 22, 1875. 

HI G. Mercalli, Vulcani in Italia, Milano, p. 116, 1883 ; and Vulcani Attivi 
della Terra, Milano. p. 264, 1907. 

•H^C. Lyell, Principles of Geology, 11th ed., vol. ii, p. 58, 1892. 

*** T. G. Bonney, Volcanoes, p. 46, 1899. 

Am. Jour. Sci. — Fourth Series, Vol. XXVII, No. 158. — February, 1909. 
10 



134 Washington — Submarine Eruptions of 1831 and 1891. 

July 8 the surface of the sea was seen to rise to a height of. 80 
feet, the column maintaining itself for ten minutes, and then 
again sinking down. This was repeated every quarter to half 
an hour, and was accompanied by a dense cloud of black smoke 
and loud rumblings. This column of smoke increased rapidly, 
and by the 12th large quantities of light-colored pumice were 
w T ashed up on the beach at Sciacca. This material was ana- 
lyzed by Abich. 

On the 16th, Giovanni Corrao saw at the base of the column 
a small island about 12 feet high with a crater in its center 
containing boiling red water. The island increased rapidly in 
size, and was visited by Professors Hoffmann and Esscher. 
They estimated its height at about 60 feet and its diameter at 
800 feet, though they do not appear to have landed. At this 
time the island had the form of a crescentic cone, with an 
opening to the sea on its southwest side, while a column of 
white smoke rose to an estimated height of 2000 feet. 

A landing was first effected on Aug. 2, by Capt. Senhouse, 
who took possession of it for the British crown and gave it the 
name of Graham Island. After this it was visited by several 
geologists and others: by Gemrnellaro on Aug. 12; by Hoff- 
mann on Sept. 26, and by Prevost on Sept. 29. 

The volcano seems to have attained its greatest dimensions 
of 65 meters high, with a circumference of about 3700 meters, 
in the early part of August, after which volcanic activity 
gradually diminished, though an eruption took place sonthwest 
of the island during this month. The general form was that 
of the summit of a cinder cone, the ridge being of uneven 
height, with a crater containing water in the center. The 
form changed materially from time to time, partly through the 
accumulation of volcanic ejections and partly through the 
destructive action of the sea, and the crater was sometimes 
breached and open to the north, and again to the south. But 
the sketches made of it are rather crude and unsatisfactory, 
for the most part. Among the best are those reproduced from 
drawings by Wright, given by Mercalli.* 

The material of which the cone was built up is unanimously 
stated to be volcanic sand, lapilli, and scoriae, arranged in 
strata partly dipping inward toward the crater and partly dip- 
ping outward on the outer portions. The absence of lava flows 
is expressly noted by nearly all the observers. Although the 
material Avas undoubtedly to a large extent scoriaceous and 
fragmentary, yet the peculiar forms shown in some of the 
sketches lead to the inference that there was involved in the 
structure an upthrust of lava, probably in a highly viscous con- 
dition and more or less broken at the top. That is, there was 
probably present an upthrust plug or spine somewhat resem- 

* Vulc. Ital., plates viii and ix. 



Washington — Submarine Eruptions of 1831 and 1891. 135 

bling, but on a much smaller scale than, that of Mont Pelee. 
In a general way, therefore, there is some reason for the belief, 
though it is now incapable of proof, that the structure of the 
island was similar to that of Bogoslof as recently described by 
Jagger,* though with by no means so perfect a spine as that 
shown in the photographs of Lacroix, Hovey, and Heilprin at 
Martinique, or in those of Jaggar. The " black rock " reported 
by Capt. Swinburnef as present in the bank remaining after 
the disappearance of the island is also evidence in favor of the 
existence of such a plug. This was 26 fathoms in diameter 
and projected to within 9 feet of the surface, and was sur- 
rounded by a bank of blocks and loose sand, sloping steeply 
down to great depths about 60 fathoms from the central rock. 

According to Davy, carbon dioxide was the only gas evolved 
in considerable quantity, but other observers report hydrogen 
sulphide and sulphur dioxide, the odor of which was perceived 
at Sciacca, so that these may be assumed to have been present. 

With diminishing volcanic activity the erosive action of the 
waves rapidly destroyed the island. By the end of October 
nearly all traces of the crater had disappeared, and the island 
was nearly level with the sea, except for a small hill of scoria 
which rose to a height of 190 feet on one side. During 
November destruction was rapid, and by the end of December 
only a small rock projected above the sea, and this soon disap- 
peared, leaving a small but dangerous shoal, with a small area 
only two and a half fathoms deep, as now marked on the 
British charts. 

An eruption near the site of Graham Island is said to have 
taken place in 16324 an( ^ another submarine volcano is men- 
tioned as having been formed in 1801. § Subsequent to the 
eruption which has been described, another is said to have 
taken place on the site of that of 1831, beginning on August 
12, 1863,|| which is reported to have formed an island about 
three-quarters of a mile in circumference, 200 or more feet 
high, and with a crater 30-40 yards across. The material of 
this is stated to have been loose ashes and sand, but none of it 
seems to have been preserved, and the island soon disappeared. 

Mercalli^f also mentions two other eruptions off the Sicilian 
coast. That of June 18, 1845, took place in lat. 36° W 56" 
K and long. 13° 44/ 36" E., near the Madrepore Bank. About 
9.30 p. m. the English vessel "Victory" felt a violent shock, 

* T. A. Jaggar, Mass. Inst. Tech. Quart., vol. x, p. 31, 1908. 

f Cf. Lyell, op. cit., p. 62. 

% Mercalli, Vulc. Ital., p. 117. 

§ E. Reclus, New Physical Geography, New York, 1886, vol. i, p. 408. 

|| Cf . this Journal (2), vol. xxxvii, 1864, p. 442 ; also Mercalli, Vulc. 
Ital., p. 120, and Reclus, op. cit., p. 408. There seems to be some uncer- 
tainty as to the exact location of this eruption. 

IT Mercalli, Vulc. Ital., p. 120. 



136 Washington — Submarine Eruptions of 1831 and 1891. 

which broke both her masts, the air was filled with a sulphur- 
ous odor, and three immense globes of fire were seen. The 
other eruption took place in the night of October 4-5, 1846, 
a little west of Girgenti, and nine miles from the coast. There 
does not appear to have been any shock, but the captain of a 
passing vessel, seeing a bright light, approached, and saw an 
immense mass of flames and smoke rising from the sea, from 
which were hurled incandescent globes. It is reported that 
the area covered by the flames was more than a mile in cir- 
cumference and that the sea appeared to be boiling. 

Petrography . 

Although it is commonly thought that the lavas of Graham 
Island were wholly basaltic, yet the descriptions of Gemmellaro 
and an analysis by Abich, as well as a brief description of one 
specimen by Foerstner, indicate that there was considerable 
diversity among them, those belonging to the earlier phases of 
the eruption having been distinctly trachytic in character. 

The specimen examined by me is in the collection of the 
Peabody Museum at Yale University, having formed part of 
an old collection of rocks, though all record of its acquisition 
has been lost, according to Professor E. S. Dana, to whom I 
am deeply indebted for his kindness in supplying the material 
for analysis and microscopical study. The small specimen is 
accompanied by a label, on which is written in now faded ink 
and in an old-fashioned script: "Lava from the volcano that 
rose from under the sea off Sciacca, 40 miles from Sicily, Aug. 
1831. Water 600 feet deep. Brought home by the Kev'd 
Eli Smith " (no date). There is no reason for doubting that it 
is what it purports to be, especially as its characters agree with 
Foerstner's descriptions of well-authenticated specimens. 

The specimens examined by Foerstner came respectively 
from the museums at Palermo, Naples, and Strasburg; the 
first having been collected by C. Gemmellaro and the last 
(which differed much from the others) being labeled as 
"lapilH floating on the sea," and thus probably belonging to 
the early period when masses of such lapilli were washed 
ashore near Sciacca, as pointed out by Foerstner. Of these 
specimens Foerstner gives an analysis only of the first. 

Megascopic. — My specimen is jet-black in color, and but 
slightly vesicular, the vesicles being very minute, and much of 
the small piece being quite compact. Foerstner's specimens 
would seem to have been much more scoriaceous and vesicular. 
Delicate, very small, glistening gray tables of feldspar are 
abundant, but no other phenocrysts are visible. 

Microscopic. — The largest and most abundant phenoci^ysts 
are of labradorite, in tables from 0*5 to 2'0 mm long, by 0*02 -0-10 
thick, which are twinned according to the Carlsbad and albite 
laws. The extinction angles show that the composition is 



Washington — Submarine Eruptions of 1831 and 1891. 137 

about Ab,Au 9 or Ab 1 An 3 . These feldspars are highly euhe- 
dral and the laths show well-formed terminal planes. The 
tables sometimes surround portions of olivine and augite phe- 
uocrysts, but carry few inclusions, mostly of black dust. Oli- 
vine is rather abundant, in subhedral, and often highly euhedral, 
equant individuals, up to 0*5 mtn in diameter. It is quite fresh 
and colorless, and carries few inclusions of ores and dust. 
Augite is rather less abundant than the olivine, in anhedral to 
subhedral, equant individuals, up to 05 mm , or rather larger. 
It is almost colorless, but slightly yellowish or greenish, and 
contains few inclusions. Opaque grains of magnetite, presum- 
ably titaniferous, are not uncommon, though small in size. 

Under low powers the groundmass is black and quite opaque, 
but high powers resolve it into a colorless, isotropic glass, 
thickly sprinkled with very minute black dusty grains. In 
general characters my specimen resembles that of Foerstner 
which he obtained from Gemmellaro, though the glass in his 
was coffee-brown. 

Chemical composition. — The results of my analysis are shown 
in column I below. The analyses of Foerstner and Abich are 
given also in II and III. 

Analysis I is, in general terms, that of an ordinary feldspar- 
basalt, and is remarkable in itself chiefly for the rather high 
Ti0 2 and the presence of considerable nickel. The high 
amount of FeO as compared with the Fe 2 3 is also a note- 
worthy feature, which will be referred to again. 

Foerstner's analysis (II) corresponds in general very well 
with mine, showing but slightly more Si0 2 and Fe a 8 and a 
little less Na 2 and K 2 0, while the differences in the figures 
for FeO, MgO, and CaO are somewhat larger, but sufficiently 
close to indicate the practical chemical identity of the two 
specimens.. His much higher figure for Al 2 3 is, of course, to 
be ascribed to the non-determination of Ti0 2 and P 2 5 , both of 
which would be weighed with the alumina. In a later paper,* 
discussing the lava of the submarine eruption of 1891 in the 
analysis of which he determined Ti0 2 , he considers that in his 
analyses of the Graham Island and Pantelleria basalts the Ti0 2 
was weighed with the silica, which should be corrected for 
about 5 per cent of titanium dioxide. That this supposition is 
incorrect is shown by the figures in analyses I and II above, as 
by the facts of analysis. The residue left on correction of the 
silica for impurities by evaporation with sulphuric and hydro- 
fluoric acids consists largely of alumina and ferric oxide, and 
does not contain all the titanium. f It may be noted in this 
connection that this residue in my analysis amounted to only 
1*73 per cent. 

* H. Foerstner, Tsch. Min. Petr. Mitth., vol. xii, p. 520, 1891. 

f Cf. W. F. Hillebrand, Bull. U. S. Geol. Surv. ; No. 305, p. 80, 1907. 



138 Washington — /Submarine Eruptions of 1831 and 1891, 





I 


II 


III 


IV 


la 


SiO„ 


48-97 


49-24 


51-87* 


61-08 


816 


A1 9 3 


16-37 


19-06 


15-30 


17-37 


160 


Fe 8 O s 


1-33 


1-77 





7-77 


008 


FeO 


8-56 


10-33 


11-40 





119 


MgO 


6-22 


5-00 


8-66 


4-02 


156 


CaO 


7'49 


8-75 


7*46 


1*46 


134 


Na 2 


4-09 


3-89 


3-90 


2-85 


066 


K 2 


1-72 


1-19 


0-85 


1-82 


•018 


H 2 + 
H 2 0- 


0-38 | 
0-08 j 


0-63 





1'63-f 




Ti0 2 


3-95 








1-45* 


•049 


P.O. 


1-04 











•007 


MnO 


0-06 





0-60 


0-62 


001 


MO 


0*08 











•001 




100-34 


99-86 


100-04 


100-09 





* "Kieselerde mit Titansaiire. *' 

f " H 2 + CI + H 2 S " = loss on ignition. 

I. Andose (feldspar-basalt scoria). Graham Island. Wash- 
ington analyst. 

II. Andose (feldspar-basalt scoria). Graham Island. H. Foerst- 
ner analyst. Tsch. Min. Petr. Mitth., vol. v, p. 391, 1883. 

III. Black basaltic scoria. Graham Island. H. Abich analyst. 
Vulk. Ersch., p. 74, 1841. 

IV. Light gray trachytic pumice. Shore near Sciacca, floated 
from Graham Island. H. Abich, analyst. Vulk. Ersch., Table 
III, No. 3, 1841. 

la. Molecular ratios of I. 

The analysis of basalt by Abich (II), made by fusion with 
barium carbonate, is very satisfactory, considering its date and 
the crude methods and reagents available at the time. It 
shows the broad chemical characters of the rock quite clearly, 
except the large amount of titanium, though the recognition 
of the presence of this constituent is noteworthy and is indic- 
ative of Abich' s acumen and the accuracy of his analytical 
work. Analysis IV will be discussed later. 

Classification. — Analysis I yields the following norm : 



Or 


10-01 ) 


Ab 


34-58 [ 


An 


21-13 ) 


Di 


7-91 I 
14-44 f 


01 


Mt 


1-86 | 


11 


7-45 C 


Ap 


2-35 


Rest 0-60 



22-35 j 

9-3i r 

2-35 ) 



65-72 



34-01 



Sal 



1-93 



Fern 
F 

L = °° 
K 2 Q+Na 2 Q 
. CaO' 



= 0-90 



KO 



100*3' 



Na 2 



= 0-27 



Class II 
dosalane. 

Order 5 
gerinanare. 
Rang 3, 
andase. 



Subrang 4 y 
andose. 



Washington — Submarine Eruptions of 1831 and 1891. 139 

The feldspar basalt scoria therefore is an an.dose (IL5.3.4), 
and it may be noted that calculation of the norm of Foerstner's 
analysis also leads to the same result. As the rock is highly 
vitreous, its mode is indeterminate. 

Variation in the magma. — The second analysis by Abich 
(IV) was made of the light-colored pumice which was washed 
up on the shore near Sciacca during the first days of the erup- 
tion, while the other (III) was made of the scoria which formed 
the island itself and which dates from the later period of the 
eruption. They are of special interest, because, in spite of their 
crudities, they indicate that the material was not uniform in 
chemical character, but that the magma underwent a change in 
chemical composition during the progress of the eruption. 

Abich, of course, does not give any description of the micro- 
scopic characters, and Foerstner's description of the light gray 
pumice (p. 392) which he also assumes to be that of the early 
phase, shows rather indefinite characters. According to him it is 
composed of flakes made up of minute, mostly colorless crystals, 
which have little action on polarized light, with redder, rust- 
brown particles and magnetite grains. The gray pumice 
contains fragments of material resembling the black basalt 
which he analyzed. He expresses the opinion that the light 
color may be due to the action of acid .vapors. But his descrip- 
tion, taken in connection with the analysis by Abich, which 
shows a marked variation in the alkalis, much lower lime,, 
magnesia and iron oxides, and higher silica and alumina, as 
compared with the basalt, indicates that this light-gray rock is 
a pumiceous trachyte, presumably described from a differen- 
tiation of the magma. 

The marked change in composition is commented on by 
Abich (p. 74), who suggests as a cause, either a gradual change 
in the depths from a silica-rich rock to one lower in silica with 
the addition of much lime (a vague forecast of the modern 
notion of differentiation), or the action of the volcanic, highly 
heated gases on pre-existing rocks, which gave rise to the gray 
pumice, his opinion favoring this latter interpretation. 

In view of the short duration of the eruption the fact of 
such a decided change in magmatic character as is indicated by 
Abich's analyses is very striking. But its greatest interest lies 
in that the change is remarkably like that shown in the succes- 
sion of magmas on the neighboring island of Pantelleria, where 
the earlier eruptions furnished highly salic and alkalic trachytes, 
rhyolites, and pantellerites, followed at the close by basalts 
very similar to that of Graham Island and the eruption of 1891. 
This would indicate for both localities a similar original magma 
and a similar course of differentiation. This matter will be 
discussed at greater length in a forthcoming paper on the rocks 
of Pantelleria. 



IK) Washington — Submarine Eruptions of 1831 and 1891. 

FOERSTNER VOLCANO, 1891. 

Bibliography. 

The eruption of 189 L, near Pantelleria, is less well-known 
than that of 1831. Lasting only about a week, it was not seen 
by any scientific observer, and the descriptions have been 
derived from the testimony of fishermen. The fullest account 
is that of A. Ricco.* Mr. G. W. Butler visited Pantelleria 
about one month after the eruption and communicates some 
notes of his own,f with a brief description of the bombs (accom- 
panied by an analysis by Gr. H. Perry). He gives a translation 
of Kicco's report,:): on which the descriptions given by Geikie§ 
and MercalliJ are also based. H. Foerstner, in a paper describing 
the rocks, 1 ! gives a brief account of the eruption, based partly 
on newspaper accounts and partly on Picco's report, though 
there are some notable discrepancies. 

This submarine volcano or eruptive center, which apparently 
gave rise to no island which projected above sea-level, has 
never been named, and is usually referred to as the submarine 
eruption of 1891, near Pantelleria. For convenience of refer- 
ence 1 would propose that it be called " Foerstner Volcano," 
in honor of the able investigator of the rocks of Pantelleria, 
adjacent to which the eruption took place, and the author of 
the best analysis and only detailed petrographic description of 
the rocks thrown out by the eruption yet published. 

The Eruption. 

The following brief account is based chiefly on Butler's 
translation of Picco's report, as the original was not accessible 
to me. During 1890 there were premonitory symptoms on 
Pantelleria, shown by increased activity of fumaroles, earth- 
quake shocks (which cracked cisterns), and the elevation of a 
part of the northeast coast. A sharp earthquake was felt during 
the night of October 14-15, 1891, when a further rise of the 
same coast line occurred, making the total elevation about 80 
centimeters, as shown by a line of white incrustations marking 
the old sea level and by the testimony of the sea-faring popu- 
lation. 

The eruption was immediately preceded by strong, sussulta- 
tory shocks on Pantelleria during the night of October 16-17, 
and the eruption began on the morning of the 17th, after which 

* A. Ricco, Comptes Rendus, Nov. 25, 1891 : and Annali Uff. Centr. 
Meteor, e Geodinam. (2), pt. 3, vol. xi, 1892. 

f G. W. Butler, Nature, vol. xlv, pp. 154, 251, 1891. 

% G. W. Butler, Nature, vol. xlv, p. 584, 1891. 

§ A. Geikie, Textbook of Geology, I, p. 334, 1903. 

If G. Mercalli, Vulcani Attivi delia Terra, Milano, p. 265, 1907. 

1[ H. Foerstner, Tsch. Min. Pet. Mitth., vol. xii, p*. 510, 1891. 



Washington — Submarine Eruptions of 1831 and 1891. 141 

the earthquakes on Fantelleria rapidly diminished and then 
ceased. Columns of * s smoke," accompanied by deep rumblings, 
were seen rising from the water about 4 kilometers west of the 
town of Fantelleria, at the northwest end of the island. Those 
who visited the spot found black, scoriaceous bombs, with much 
steam, rising to the surface along a narrow line about 850- 
1000 meters long, and directed northeast-southwest. Both 
Ricco and Butler expressly state that no solid island was formed, 
but that the erupted material was solely in the form of floating 
bombs. According to Foerstner, on the other hand, by October 
18, when the eruption seems to have been at its height, there 
was formed an island about 1000 meters long, 200 meters wide, 
and 10 meters high. The existence of a stable island may, 
however, be doubted, in spite of these detailed figures, in view 
of the explicit denial of Ricco and Butler. 

Continuing their description of the eruption, vast numbers 
of subspherical bombs rose to the surface, the largest having a 
diameter of more than one meter. Some were thrown to a 
height of 20 meters, and many ran hissing over the water, * 
discharging steam, and being kept afloat hy the gases con- 
tained in their vesicles. The bombs were hot, some when 
recovered sufficiently so to melt zinc (420° C), and a few were 
red-hot in daylight. After floating for a time, most of the bombs 
exploded, the explosions succeeding each other so rapidly as to 
resemble the noise of a battle, and the fragments sank to the 
bottom. An odor "as of gunpowder" was noted, and some 
H 2 S and SO., seem to have been emitted. The eruption 
ceased on October 25, and little change appears to have been 
produced in the sea bottom (which, however, w T as not well 
known previously), though the present charts show a small 
area, of only about 30 fathoms, at the site of the eruption. 

The most striking feature of this eruption, the ejection of 
the material in the form of ellipsoidal or spheroidal masses, 
called " bombs," is of interest as bearing on the origin of the 
so-called " pillow " lavas, which consist of aggregates of such 
forms. A very complete bibliography of these up to 1899 is 
given by J. Morgan Clements, f who compares the ellipsoids 
of the Crystal Falls District with the block lavas of Giorgios 
Ivaimeni at Santorini. Among later papers treating of the 
subject may be mentioned those by I. C. Russell, J R. A. 
Daly,§ and Reid and Dewey. || 

The origin of the structure is variously explained, a few 

*One is reminded of the behavior of a globule of metallic sodium on 
water. 

fMon. U. S. Geol. Surv., xxxvi, p. 112, 1899. 
t Bull. II. S. Geol. Surv., No. 199, p. 113, 1902. 
£ Amer. Geol., vol. xxxii, p." 74, 1903. 
|| Quart. Jour. Geol. Soc, vol. lxiv, p. 267, 1908. 



1^2 Washington — Submarine Eruptions of 1831 and 1891. 

attributing it to the accumulation of bombs or the viscosity of 
pahoehoe lava, ejected subaerially, while the greater number 
connect the structure with eruption under submarine condi- 
tions or with intrusions into mud or silt. The authors of the 
last three papers cited, as well as Geikie and Teall, attribute it 
to truly submarine eruptions, and in the discussion of the 
latest paper Dr. Flett* cites "bombs " of the 1891 eruption in 
analogy with the ellipsoids of Cornwall. It is noteworthy 
that, according to Butler's descriptions, the " bombs " of this 
agree in many respects with the Cornwall masses, especially 
in their highly vesicular texture, arranged in bands of some- 
what varying characters. As according to the account of 
Ricco, some of the "bombs" of the Foerstner Yolcano were 
red-hot, even after floating for some time, it is evident that 
they must have been distinctly viscous, so that those which 
did not reach the surface and were piled up on one another 
must have been more or less flattened and distorted by the 
superincumbent mass, and thus have given rise to the pecu- 
liarities of shape and mutual fitting together of the curved 
surfaces so well shown in the illustrations to. the papers cited 
above. 

Some experiments by Johnston-Lavis, who found that 
globules were produced by injecting one highly viscous liquid 
into another, are cited by G. Plataniaf as illustrating the 
formation of the spheroidal basalts of Aci-Castello. Such a 
division into spheroids would be favored were the injected 
material molten rock, when the influence of steam and the 
spheroidal state of water, or rather here of the injected lava 
surrounded by a layer of steam, would come into play, 
especially if the depth, and hence the supply, of water was so 
great as to prevent its exhaustion by evaporation. 

It is notable in this connection that the examples of pillow- 
lavas seem to be wholly confined to the so-called " basic " rocks, 
those low in silica and high in femic constituents : whereas w r e 
find at eruptions in the open sea of lavas rich in silica (as at 
Santorini) that the mass has solidified into angular blocks. 
The difference is to be ascribed to the difference in the relative 
fusibility and viscosity of the two kinds, the highly femic 
magmas lending themselves readily to the assumption of a 
more or less spherical form as the mass separates on issuing 
into the water, while the very viscous siliceous magmas would, 
on comparatively small cooling, become so nearly solid as to 
crack, and their rapidly complete solidification would prevent 
the rounding of the sharp angles of the blocks. 

* Quart. Jour, Geol. Soc, vol. lxiv, p. 270, 1908. 

fin H. J. Johnston-Lavis, the South Italian Volcanoes, Naples, p. 42, 1891. 



Washington — Submarine Eruptions of 1831 and 1891. 143 

Although the eruption of Foerstner Volcano ceased in 
October, 1891, there appears to have been another submarine 
eruption in December of the same year, which took place 
south of Pantelleria and is said to have formed a small island 
about 500 meters in diameter.* Foerstner also mentions the 
formation of a small island " in the waters of Pantelleria " in 
July, 1881, observed by Captain Swinburne, but he cites no 
authority and there would seem to be some error in the date. 
This is not mentioned by Fuchsf in his list of volcanic erup- 
tions of 1881, nor by Mercalli. 

Petrography. 

The petrographical characters of the bombs have been briefly 
described by Butler,^ and in greater detail by Foerstner.g 
The material examined by me was obtained in 1897 from Mr. 
Francis H. Butler, of London, a cousin of the author of the 
papers cited above, who collected with him on Pantelleria. 
Mr. Butler wrote me as follows in a letter dated 21st Novem- 
ber, 1907, by which it appears that there can be no doubt as to 
the authenticity of my specimen : 

" My specimens of the Pantelleria rocks were collected by 
myself early in November, 1891. The submarine lava was 
floated to the sea among the fishing boats N.E. {sic) of Pan- 
telleria during an eruption in, I think, October, 1891 

The vesicular submarine lava from which my specimen came 
was obtained by some fishermen. Many of the masses of lava 
floated to the surface exploded, and all, after awhile, that were 
not saved by the fishermen became waterlogged and sank. I 
was present at the purchase of the big mass of submarine lava 
from which your specimen came : the greater part of the block 
was presented to Prof. Judd for the Poyal College of Science." 

According to Gr. W. Butler the bombs show a brownish out- 
side layer, about one inch thick, due to the vesiculation of 
brown glass, which contains phenocrysts of triclinic feldspar, 
olivine, magnetite, and probably augite. Beneath this is a 
darker layer about one-half an inch thick, which is mostly 
glass with the same minerals. The greater part of the bomb 
is coarsely spongy, perfectly black with a pitchstone-like luster, 
and more highly crystalline in thin section, composed of about 
one-third triclinic feldspar, olivine, and augite, and the rest a 
black groundmass, opaque in thin section. 

My specimen probably came from the interior, as may be 
inferred from this description and from the small amount of 

*H. Foerstner, Tsch. Min. Petr. Mitth., vol. xii, p. 510, 1891. 
fC. W. C. Fuchs, Tsch. Min. Petr. Mitth., vol. v, pp. 97-101, 1883. 
X G. W. Butler, op. cit., p. 251. 
£ H. Foerstner, op. cit. , p. 513-518. 



144 Washington — Submarine .Eruptions of 1831 and 1891. 

salt extracted by leaching. It is coaly black, with a slightly 
brownish tinge and pitchy luster ; very highly vesicular, with 
small, rounded vesicles, the size of which varies somewhat in 
different parts of the specimen, as is also true of Foerstner's 
piece. No phenocrysts are visible. 

Microscopic. — Judging from Foerstner's description and the 
examination of my specimen there is considerable variation in 
the microscopic character, though not more than is usual in 
such glassy scorias. The most abundant and prominent pheno- 
crysts are of a lime-soda feldspar, showing Carlsbad and less 
often albite twinning, the extinction angles of which indicate 
the composition Ab 1 An 2 , which is somewhat more sodic than 
appears to be true of Foerstner's rock. These feldspars are 
tabular and very thin in my specimen, measuring only up to 
0.5mm j on g !^ T ^ ot more t | 1 ' au o-Ol — 0-05 mm thick; while in 
Foerstner's they would appear to be larger and more euhedral, 
and more nearly like those in my specimen of the Graham 
Island lava. The phenocrysts of augite are few, up to about 
l mm in diameter, subhedral and stoutly prismatic, of a brownish 
or greenish gray. Colorless olivine is present, apparently in 
less amount than the augite and in smaller, mostly anhedral 
grains, though a few show euhedral rhombic sections. 

The groundmass in my specimen consists of a brownish, 
clear glass, which is very thickly sprinkled with a rusty brown 
or black dust, so much so as to be almost or quite opaque in 
places. Foerstner notes two varieties of the glassy ground- 
mass ; one highly vitreous, and the other less so, in which the 
glass is present in small amount as a cement or mesostasis. He 
describes these in great detail, which it is needless to quote 
here. 

Chemical composition. — The material of the bombs ejected 
by Foerstner Volcano has been analyzed by Foerstner and by 
Perry, the results being given in II and III below. A new 
analysis made by me is given in I, the rock powder having 
been leached with water, as in the preceding case, by which 
process only a very small amount of salt was extracted. Foerst- 
ner states that his specimen contained about 25 per cent of sea 
salts, so that his probably came from the exterior, and mine 
from the interior, of a bomb. 

My analysis shows low silica and alumina, very low ferric 
oxide but very high ferrous oxide, low magnesia and alkalis 
and rather high lime and phosphorus pentoxide. The most 
striking feature is the very large amount of titanium dioxide, 
which is almost unparalleled elsewhere among volcanic rocks, 
and is only exceeded in the melilite-basalts of the Hegau.* 

* U. Grubenmann, Inaug. Diss. Zurich, 1886. 



Washington — Submarine Emotions of 1831 and 1891. 145 



SiO„..-. _.. 44-83 

A1 2 3 

Fel0 3 

FeO .. 

MgO 

CaO 

Na„0 

H 2 + 

H o 0- 

k;o 

Ti0 2 

PA--- 

MnO 



I 


II 


III 


la 


44-83 


44-64 


46-40 


747 


11-73 


12-74 


21'84 


115 


1-35 


4-21 


9-53 


009 


11-79 


11-17 


2-04 


164 


5-50 


5-82 


5*37 


•138 


9'63 


10-12 


10-33 


•172 


3-34 


4-31 


3-27 


•054 


0-81 ) 
0-10 j 


0-51 


n.d. 










1-40 


1-41 


1-69 


•015 


6-88 


5-86 


n.d. 


•086 


214 


n.d. 


n.d. 


•015 


0-20 


0-20 


n.d. 


•003 


99-70 


100-99 


100-47 





Analyses of basalt scoria from Foerstner Volcano (1891) : 

I. "Analysis by H. S. Washington. 
II. Analysis by H. Foerstner, Tsch. Min. Petr. Mitth., xii, 
p. 512, 1891. 

III. Analysis by G. H. Perry, Nature, vol. xlv, p. 252, 1891. 
la. Molecular ratios of I. 

The amount of nickel was not determined, but it is probably 
about the same as that found in my analysis of the Graham 
Island basalt (about 0'10 per cent), as shown by the depth of 
the greenish tint of the filtrate from the ammonia precipitate. 

Foerstner' s analysis (II) agrees very closely with mine in 
most respects, and especially shows about the same amount of 
Ti0 2 and a similar very high ratio of ferrous to ferric oxide, 
though his figure for the latter is decidedly higher than in I. 
His alumina must be corrected for one or two per cent of P 2 5 . 

The alumina of Perry's analysis (III) must be corrected for 
about 9 per cent of Ti0 2 and P 2 5 , the subtraction of which 
would bring the figure for alumina close to those of the other 
analyses. The figures for the iron oxides are reversed as com- 
pared with those of Foerstner and myself, and must be 
regarded as incorrect, especially as Perry himself expresses 
doubt as to their accuracy.* The somewhat higher silica may 
be due to its not having been corrected for impurities by 
evaporation with HF. The other figures are closely concord- 
ant with those of the other two analyses. 

Classification. — The norm of I is calculated to be as 
follows : 

* He remarks, "As the powder was magnetic, Fe 2 3 was probably com- 
bined with FeO to form Fe 3 4 . This would give FeO 7*55, Fe 2 3 " 4'18." 
The basis of calculation for these fig-ares is not clear ; as they yield a molecu- 
lar ratio of FeO : Fe 2 3 =4:1. 



146 Washington — Submarine Eruptions of 1831 and 1891. 

Or 8-34 ) Sal ni . TTT 

Ab 28-30}- 49'43^- =1*00 • <- J ass 111, 

An 12-79 \ ^m salfemane. 

Di 17-22 i _ ._ Order 5, 

Hy 5-21 [• 29-21 

Ol _ 6-78 



L gallare. 
K 2 + Na 2 

II 13-07 j" i0iD | CaO' " camptonase. 

Ap 5-04 5-04 J KO a x. 

Rest .... 0-91 — ^- = 0-28 Subrang 4, 
Na O camptonose 



Mt 2-09 [ J, 49-41 -— 2 -— *— _^_ = 1.50 Ranges, 



• According to my analysis, therefore, the rock is a camptonose 
(III. 5. 3. 4), and falls almost exactly at the center of all the 
divisions, except that it tends to be domalkalic. As regards 
the femic Constituents, which may be briefly considered 
because the rock is salfemic, it falls in the dopyric section of 
the dopolic grad, and in the premiric section of the permirlic 
subgrad, as brought out by the following ratios : 

^ ,, P + ° o \ ' * a Di + Hy 

Grad - - = 1-93, Section of grad J ~ = 3-31. 

M Ol 



Subgrad 



(Mg, Fe)0 + CaO" 

= GO 



Na 2 0" 

(Ms Fe)0 

Section of subgrad ^-^ — — 2*35. 

to CaO" 

Mt 
The rock is also prehemic, since -—- = 0-16, but no provision 

has been formally made for recognition of this relation in 
rocks of the salfemane class, though the relations of Fe 2 3 and 
Ti0 2 are expressed in the suborders of classes IY and Y.* 

According to Foerstner's analysis, as given by him, the rock 
falls in monchiquose (III. 6. 2.4), because of the higher soda 
and the considerable amount of lime that enters normative 
anorthite and diopside, and thus takes up so much silica as 
to give rise to the presence of much nephelite in the norm. 
If, however, the alumina is corrected for two per cent of 
P 2 5 , the subrang becomes kilauose (III. 5. 2.4), since much 
of the lime is thus used in forming apatite, leaving more 
silica for the soda. The relations are shown below. 

These relations, while of no importance in themselves, are 
of interest in showing the necessity of the determination of the 
minor constitutents for the purposes of the quantitative classi- 
fication. The norm of III need not be considered. 

*Cf. Cross, Iddings, Pirsson, and Washington, Quant. Class, p. 134, 1903. 



Washington — Submarine Eruptions of 1831 and 1891. 147 



Norms of II. 



Corrected 
8-34 

27-25 
5-84 
4-83 

24-50 
7-44 
6-03 

11-10 
5-00 

In the prevalent classifications the rock is a vitreous feld- 
spar-basalt scoria. The mode need not be discussed, as it is 
indeterminate, owing to the large amount of glass present. 

Correlation of the Basalts. 
In the following table are presented a number of analyses 
made by me of basalts from the same general region and of. 
characters similar to those here described. 



Uncorrected 

Or 8-34 

Ab 16-24 

An 11-40 

Ne_ 10-79 

Di 31-30 

01 5-02 

Mt 6-03 

II 11-10 

Ap none 



I 


II 


III 


IV 


V 


VI 


VII 


VIII 


48-97 


44-83 


45-72 


46-22 


46*55 


48-84 


47-66 


43-64 


16-37 


11-73 


12-45 


12-23 


14-55 


14-62 


14-36 


13-52 


1-33 


1-35 


1-57 


4-91 


3-17 


2-08 


2-33 


6-40 


8-56 


11-79 


12-01 


7-71 


7-88 


9-00 


8-44 


5-52 


6-22 


5.50 


5-29 


6-74 


8-61 


7-15 


8-19 


9-36 


7-49 


9-63 


9-58 


9-86 


8-75 


9-33 


9-36 


9-52 


4-09 


3-34 


3-40 


3-39 


3-71 


2-86 


3-51 


3-89 


1-72 


1-40 


1-08 


1-13 


1-62 


0-89 


1-54 


2-18 


0-38 


0-81 


0-40 


0-17 


0-14 


0-49 


0-17 


0-49 


0-08 


o-io 


o-oi 


0-05 


0-03 


0-07 


0-20 


0-16 


3-95 


6-88 


6-43 


5-68 


3-84 


3-57 


3-83 


4-55 


1-04 


2-14 


1-54 


1-46 


'0-55 


0-36 


0-45 


0-74 


0-06 


0-20 


0-16 




o-io 


0-04 







0-08 




0-15 





0-12 


0-08 


SrOO-03 



Si0 2 

A1 2 3 

Fe 2 O s 

FeO 

MgO 

CaO 

Na 2 

K 2 

H 2 + 

H 2 0- 

Ti0 2 

PA 

MnO 
NiO 



100-34 99-70 99-82 99'55 99*62 99*89 100'54 99*60 

I. Andose (II.5.3.4). Graham Island, 1831. 

II. Camptonose (III.5.3.4). Foerstner Volcano, 1891. 

III. Camptonose (III.5.3.4). Dike in Costa Zeneti, Pantelleria. 
Quart. Jour. Geol. Soc., lxiii, p. 74, 1907. 

IY. Camptonose (III.5.3.4). Monte Sant' Elmo, Pantelleria. 
Quart. Jour. Geol. Soc, lxiii, p. 74, 1907. 

V. Camptonose (III.5.3.4). Block in tuff, II Fosso, Linosa. 
Jour. Geol., xvi, p. 23, 1908. 



VI. Auvergnose (III.5.4.4-5). 
Geol., xvi, p. 17, 1908. 
MVII. Camptonose (111.5,3.4). 
Jour, xxiv, p. 239, 1907. 

VIII. Monchiquose (III. 6. 2.4] 
This Jour, xxiv, p. 239, 1907. 



Monte Ponente, Linosa. Jour. 

Castellfullit, Catalonia. This 

La Garrinada, Olot, Catalonia. 



14$ Washington — Submarine Eruptions of 1831 and 1891. 

Apart from the higher silica and alumina of the earlier one, 
the basaltic lavas of the two submarine eruptions are closely 
similar, I being slightly higher in MgO and Na 3 and lower 
in FeO, CaO, and P 2 5 . The variation in TiO a is also consider- 
able, but in both rocks the amount of this constituent is higher 
than usual among basaltic rocks. The distinctly more femic 
character of the later eruption is in harmony with the decidedly 
more siliceous and salic character of the material of the early 
phase of the Graham Island eruption, as shown by Abich's 
analysis to which attention was called above, and there is thus 
indicated a progressive change in composition with increase 
in femic constituents for the general magma of the region. A 
change in the same general sense has already been noted as 
having occurred on Pantelleria, and a similar relation is observed 
on Sardinia, where the large volcanoes of Monti Ferru and Arci 
poured forth respectively phonolitic trachytes and rhyolites in 
the early stages of activity, followed by extensive outflows of 
basalt. On this island too all the small cones of the most recent 
date are basaltic, though varying somewhat in composition, 
and in general decidedly more salic than are those of Pantelleria, 
Linosa, and the two submarine eruptions. - The basalts of 
Catalonia also show great analogies with those of Pantelleria 
and the neighboring volcanic vents, as has been elsewhere 
pointed out. 

Aside from the very high Ti0 2 of all these basalts, one of 
their most striking characters is the predominance of ferrous 

over ferric oxide. This ratio _ ~ reaches a maximum in the 

. Fe A 
basalts of the two submarine eruptions and in that of the 

dike on Pantelleria, which are highly vesicular, and shows 

a minimum in the also very vesicular lava of La G-arrinada, 

near Olot. Lavas showing ratios between these extremes 

are either only slightly vesicular or are solid flows. 

In a former paper* the difference observed at the Catalan 
volcanoes was ascribed to the more favorable conditions for 
oxidation of the ferrous iron when the lava issued in a highly 
vesicular condition. Since, however, we find that oxidation is 
at a minimum in the highly vesicular basalts described in the 
present paper, some discussion of the matter will be pertinent. 

The ready oxidizability of the ferrous iron in rocks on heat- 
ing in air, even when in the solid though powdered form, is 
well known to all who have had to make rock analyses, and 
the researches of Gautier f have shown that steam alone' also 
exerts a powerful oxidizing action on ferrous silicates. On the 
other hand, further researches of Gautier and of Hiittner showed 

* H. S. Washington, this Journal, vol. xxiv, p. 240, 1907. 

fCf. F. W. Clarke, Bull. U. S. Geol. Surv., No. 330, 1908, p. 228. 



Washington — Submarine Eruptions of 1831 and 1891. 149 

that C0 2 can be reduced to CO by the action of hydrogen at 
elevated temperatures. Gautier also showed that the reaction 
is reversible, at a white heat being : 

C0 2 + H 2 = CO + H 2 0, 

and at temperatures between 1200° and 1250 ° 

CO + H 2 = C0 2 + H 2 . 

As Clarke remarks : "When water emitted by heated rocks 
(or that of the sea, H. S. W.) mingles with carbon dioxide, 
within the vent of a volcano, both reactions take place, and 
mixed gases, which sometimes contain a trace of formic acid, 
are generated. This mixture is a powerful reducing agent, 
which acts upon the iron silicates in an opposite direction to that 
of the oxidizing vapor of water. Either oxidation or reduction 
is therefore possible, according to the preponderance of one 
constituent or another among the volcanic gases." 

We may therefore ascribe the difference in the iron oxide 
ratios in the scoriaceous lavas to the fact that, in one case they 
were erupted subaerially and in the other subaqueously. In 
the former large quantities of heated air would have acted 
on the highly vesicular masses during their ejection, and would 
have materially aided the steam present in oxidizing the ferrous 
iron of the magma. In the other, free atmospheric oxygen 
was wanting and, as we know that carbon dioxide was emitted 
in large amount at Graham Island, and presumably also at 
Foerstner Volcano, we may suppose this and the steam to have 
reacted with the formation of carbon monoxide and hydrogen, 
which would have had a powerful reducing effect on any ferric 
iron present in the magma. In the case of the compact and 
scarcely vesicular lavas, such as those of Catalonia, which were 
ejected in the form of massive flows, there would have been 
comparatively little opportunity for either an oxidizing or a 
reducing action, and we should expect to And, as we do, ratios 
of ferrous to ferric oxide intermediate between those of the 
scorias, and presumably more nearly like that of the original 
magma. 

The reducing conditions at the submarine eruptions must 
also have been very materially increased by the hydrogen sul- 
phide and sulphur dioxide, which we know to have been 
present. In this connection it is of interest to recall that John 
Davy found that the waters of the two small craters of Graham 
Island, on August 5, contained hyposulphite of lime and 
magnesia, in addition to sulphates, but no salts of the alkalis. 
Although the condition of chemical analysis at the time does 
not give this statement much weight, yet it may be significant. 
Abich remarks on the absence of hydrochloric acid during the 

Am. Jour. Sci.— Fourth Series, Vol. XXVII, No. 158.— February, 1909. 
11 



150 Washington — Submarine Eruptions of 1831 and 1891. 

whole progress of the Graham Island eruption, a gas which is 
so commonly observed at other volcanoes, as Vesuvius. 

The very high FeO : Fe 2 3 ratio shown by the basalt of the 
Zeneti dike does not seem to be explicable by the hypothesis 
advanced above, as it is decidedly vesicular and jet apparently 
not due to a submarine eruption. Discussion of this rock 
must, however, be reserved for a later publication, in connec- 
tion with the geology and petrography of Pantelleria. 

In both his papers cited above Foerstner suggests that the 
basalts of the eruptions of 1831 and 1891, as well as those of 
Pantelleria, are connected with those of Etna. Of the lavas 
of Etna, however, we have no modern or satisfactory chemical 
analyses, strange as it may appear. With the exception of one 
of an ash by Picciardi in 1884,* none of them are later than 
1881. The lavas and ashes of Etna have been analyzed^ by 
O. Silvestri, Fuchs, Giimbel, Von Lasaulx, and Picciardi. 
Most of the analyses show a general resemblance to each other, 
as well as to those by Foerstner of the basalts of Graham Island 
and Pantelleria, in which Ti0 2 and other minor constituents 
have not been determined. The studies of Von Waltershausen 
and Von Lasaulx also show that, while the lavas of Etna are 
predominantly basaltic, there are numerous flows and dikes of 
more salic rocks, especially belonging to the earlier phases of 
this volcano, a fact already noted by Abich. There is, there- 
fore, reason for the belief that the magmas of Etna and of 
Pantelleria and the submarine eruptions are chemically similar 
and probably genetically connected. But, as a marked charac- 
teristic of the basalts of the latter volcanoes is the high content 
in titanium, it is clear that modern, complete analyses of the 
Etna lavas are needed before the matter can be discussed intel- 
ligently. 

Locust, N. J., August, 1908. 

*L. Ricciardi, Att. Ace, Gioen., vol. xviii, p. 4, 1884. 

f J. Roth, Beitrage zur Petrographie, p. exxviii, 1869, and pp. lxxvi-lxxx^ 

1884. 



E. II Sellards — Types of Permian Insects. 151 

Art. IX. — Types of Permian Insects ; by E. H. Sellards. 

[ Continued from vol. xxiii, p. 355, May, 1907.] 
Part III. — Megasecoptera, Oryctoblattinid^e and Prot- 

ORTHOPTERA. 

The terms Dromeus, Scopus and Therates, proposed for 
genera of ephemerids in Part II of this paper, have been found 
to be preoccupied. As substitutes I suggest for Dromeus 
Sellards ( non Reiche ), Misthodotes ; for Scopus Sellards ( non 
Megerle), Mecus ; for Therates Sellards (non Latrielle), Esca. 
The term Rekter should read Recter ; while Twpus (Pt. I, 
p. 249 ) should read Typus. For these corrections I am 
indebted to Mr. Leonhard Stejneger of the National Museum. 

Megasecoptera. 

The Megasecoptera have four slender, equally developed 
wings, which are broadest at the middle and narrowed at the 
base. The anal area of the wing is notably reduced. Cross 
veins are not numerous. The abdomen is long, and is termin- 
ated by streamers. 

The group is abundantly and typically developed in the 
Commentry Coal Measure deposits of France,* and is sparingly 
represented in some other European, and in the American 
Coal Measure deposits.! The Megasecoptera have been known 
heretofore from the Carboniferous only. Their extension 
into the Permian is indicated by the genus herein described. 

Opter Urongniartii gen. et sp. n. Text-figure 7. 

The genus Opter is based upon a single detached wing, the 
apex and a part of the inner border of which are wanting. 
The wing is thin, slender, and very much narrowed at the base. 
The costal border is straight. The subcosta is a thin vein run- 
ning parallel to the radius. The media is united with the 
radius in the basal one-fourth of the wing. At its point of 
separation from the radius the media has a characteristic 
downward curve ^almost touching the cubitus. The cubitus 
is two-branched. Cross veins occur sparingly in the w T ing. 

Length of wing estimated, 13 to 15 mm ; width, 2-J to 3 mm . 

Type specimen No. 1286. 

* Charles Brongniart, Tnsectes fossiles des Temps primaires, pp. 280 et 
seq., 1893. 
f Anton Handlirsch, Die Fossilen Insekten, p. 312 et seq., 1906. 



152 F. II. Sellards — Types of Permian Insects. 

Oryctoblattinidce. 

The family Oryctoblattinidse has been erected recently by 
Handlirsch and is placed by liim among the Protoblattoidea.* 
The genera referred to this family by its author are as follows: 
From the Carboniferous ; Oryctoblattina Scudder, Blattinopsis 
Giebel, Anadyomene K. v. Fritsch, Glaphyrophlebia Hand- 
lirsch, Microblattina Scudder, Prisca K. v. Fritsch, and Phi- 
pidioptera Brongniart ; from the Permian, Oryctomylabris 
Handlirsch, and Pseudoftdgora Handlirsch. Two additional 
genera obtained from the American Permian are added in the 
present paper. 

The family contains several forms of particular interest. 
The genus Phipidioptera was originally referred by its author, 
Brongniart, directly to the Homoptera, and to the recent 
family, Fulgoridge.t The forms known in literature as Ful- 
gorina Goldenbergii Brongn. and F. ovalis Brogn., and regarded 
by some as Homoptera, are included in this revised classification 
in the genus Blattinopsis Giebel. The Permian form known 
as Fulgora Fbersi Dohrn (Fidgorina Fbersi Gold. ) stands as 
the type of the new genus Pseudofulgora Handlirsch. Some 
of the other genera, including the type genus, Oryctoblattina, 
bad been previously regarded as cockroaches. It is thus seen 
that the family is made up of forms originally referred to 
very diverse groups. The geological range of the Oryctoblat- 
tinidae as now known is through the Upper Carboniferous, and 
into the Permian. 

The wing venation in this family is characteristic. The 
numerous veins uniting the subcosta with the costa have the 
appearance of cross veins rather than of oblique branches, 
differing in this respect from the blattid wing. Subcosta and 
radius are distantly separated and are united by oblique cross 
veins. The radical sector arises early and is repeatedly sub- 
divided. The media is two to several times branched. The 
cubitus has usually numerous inferior branches. The anal 
area is marked off by a thin depressed line, and is traversed 
by a few strong veins. 

That part of the wing lying between the sector and the 
cubitus affords the specially characteristic features of the vena- 
tion. IViost genera have a plainly marked line extending 
across this area from the cubitus to, or beyond, the sector, thus 
marking off the basal from the apical half of the wing. This 
cross line is plainly marked in Blattinopsis, Anadyomene, 
Prisea, Pseudofulgora, the two Permian genera described 
below and in two undescribed genera obtained by the writer 

* Revision of American Paleozoic Insects, 705, 1906 ; Die Fossilen Insekten, 
p. 155 and p. 346, 1906. 

f Insectes fossiles, p. 445, 1893. 



E. H. Sellards — Types of Permian Insects. 153 

from the Kansas Coal Measures. The line does not appear in 
the illustrations of Oryctoblattina, Glaphyrophlebia, Rhipidiop- 
tera, and Oryctomylabris, although the venation is otherwise 
closely similar. The cross line marks the location of a more 
or less distinct break in the continuity of the veins. The 
apical part of the wing beyond the line is closely filled with 
veins. 

Among modern insects a similar interruption in the vena- 
tion by a cross line is best developed and most often observed 
in Hemiptera, and has led, perhaps not without reason, to the 
comparison of some members of this family with Hemiptera. 
In pointing out recently the existence of a similar line in the 
wing of Cicadidse, Woodworth writes* : " The most curious 
feature of this venation is a mark extending across the wing, 
which can be seen only in certain lights upon the membrane; but 
wherever this line crosses a vein, it is very evident, because 
the vein is here entirely interrupted ; " and p. 123, "In this 
connection a similar structure in the fossil Fulgorina is, as 
already pointed out, of interest. " 

If this peculiarity in the wing venation occurred among 
Hemiptera only (Homoptera and Heteroptera) the argument 
for the Hemipterons affinity of the Oryctoblattinidse would be 
very strong indeed. Handlirsch states, however, that a similar 
break occurs in some Mantidse and in many other insects (p. 
158, Die Fossilen Insekten, 1906.) 

The body structure is unfortunately still very imperfectly 
known. Brongniart states that certain specimens representing 
this family from the French Coal Measures have parts of the 
body preserved. JNp illustrations of these specimens are given 
and the description is brief. The body is said to be short ; 
the head rather large ; the eyes large, round, and salient. The 
antennae are described as in some specimens long, in others 
short. Regarding the mouth parts, it is stated that two little 
pieces are observed in front of one of the impressions which may 
be mandibles. f The hind wings of the family are unknown. 

Until more definite information regarding the body struc- 
ture and the hind wings of the Oryctoblattinidse is obtained a 
satisfactory placing of the family seems hardly possible. 

Pursa ovata gen. et sp. n. Text figure 4. 

This is a genus of small Oryctoblattinidse. The wings are 
arched, the apices bluntly rounded. The radial sector arises 
as a thin vein near the termination of the basal third of the 
wing. The media is broken up at the base. Its attachment 

*The Wing Veins of Insects ; by C. W. Woodworth, Univ. of California, 
Technical Bulletin No. 1, Agri. Experiment Station, p. 122, 1906. 
f Insectes fossiles, p, 446, 1893. 



154 E. II. Sellards — Types of Permian Insects. 

is apparently with the sector. The cubitus has several weak, 
inferior branches. A complete interruption of the veins 
occurs at the cross line in this genus. The apical part of the 
wing is traversed by numerous simple, parallel veins. 
Length of wing, 8 mra ; width, 3 mm . Type, No. 1126. 

Sindon speciosa gen. et sp. n. Text figure 1. 

The costal border of the wing of species of this genus is 
slightly arched. The radial sector in the type species is 
divided into several branches, each of which is forked at the 
tip. The media is two branched. The cubitus has numerous 
strong arched parallel branches. The interruption of the veins 
at the middle of the wing is much less marked than in the 
genus Pursa. Cross veins are numerous in the basal half of 
the wing, but are lacking in the apical half. 

Length of wing, 8 mm ;' width, 2J mm . Type, No. 85. 

Protorthoptera. 
The Protorthoptera although more primitive are in a general 
way related to modern Orthoptera. The venation of the front 
wing, while specialized in various ways, is still of a compara- 
tively generalized type, the main veins of the wing being as a 
rule readily recognizable. The hind wing is broader than the 
front ; the anal area expanded and folded. This group of 
insects is predominant in the Kansas Permian deposits, both 
in number of species and of individuals. 

Lepiidw, family new. 

The front wings of the Lepiidse are elongate, and of 
coriaceous texture. The main veins are strong and are plainly 
impressed in the thickened membrane. The interspaces between 
the veins are filled with a net-work of areoles, the lines, when 
seen from the top surface, showing as ridges on the membrane. 
The subcosta lies close to the radius and extends to or beyond 
the middle line of the wing. The radius gives off oblique 
branches beyond the termination of the subcosta. The sector 
is free and two or three branched. Media is also free and 
several times branched. Cubitus divides early in the front 
wing, the inner branch being simple. The main branch of the 
cubitus, Cu n is directed toward the inner border, but after 
approaching the border bends abruptly forward, and extends 
parallel with the border beyond the middle line of the wing, 
giving off several inferior branches. The anal area is marked 
off by a thin line and is traversed by two simple veins. 

The hind wings are broad, the anal area expanded and folded. 
The folded part of the wing is thin, the areoles lacking. The 



E. H. Sellards — Types of Permian Insects. 155 

part of the wing not folded is coriaceous and areolated in a 
manner similar to the front wing. The sector of the hind 
wing either is detached from the radius or joins that vein close 
to the base. Cubitus is deeply impressed and may be reduced 

Figs. 1-8. 










Explanation of Text Figures 1-8. 

Fig. 1. Sindon speciosa gen. et sp. n. Front wing. No. 85. 

Fig. 2. Horates elongatus gen. et sp. n. Front wing. No. 992. 

Fig. 3. Atava ovata gen. et sp. n. Hind wing. No. 372. 

Fig. 4. Pursa ovata gen. et sp. n. Front wing. The apex of the wing is 
restored from the counterpart of the same specimen. No. 1126. 

Fig. 5-6. Liomopterum ornatum gen. et sp. n. Front and hind wings. 
No. 5. Apex of the hind wing restored from a second specimen of this 
genus ( No. 657 ). 

Fig. 7. Opter Brongniartii gen. et sp. n. Front wing. No. 1286. 

Fig. 8. Lepium elongatum gen. et sp. n. Front wing. No. 132. 

All illustrations four times natural size. 



156 E. II. Sellards— Types of Permian Insects. 

to a simple straight vein. More commonly, however, it gives 
off a vaulted and forwardly directed branch. 

The family is described from the wings, the body being 
unknown. 

Lepium elongatum, gen. et sp. n. Text figure No. 8. 

The wings of the insects of this genus are elongate with 
rounded apex and narrowed base. The radial sector is given 
off not far from the termination of the basal third of the wing,, 
and in the type species is two branched. Media is free. It 
divides early, M, being deeply forked. Near the termination 
of M 2 the areoles pass into accessory veins, of which there are 
four. From Cu, several (about eight) inferior veins pass to the 
border. The anal area is traversed by two simple veins. 

Length of front wing, 12 mm ; width, 4 mm . Type, No. 132. 

Lepium reticulatum sp. n. 

This species is described from one complete front wing with 
which is associated parts of the hind wing. Radius reaches 
farther along the costal border toward the apex than does the 
radius of L. elongatum,. The sector is three instead of two 
branched, one branch being lost, however, in the areoles of the 
wing, and not reaching the border. Media is similar to that of 
the type species, as are also the cubital and anal areas. 

The anal area of the hind wing is expanded, thin, and plicated. 
Areoles in this part of the wing are lacking. Of the remainder 
of the hind wing the apical and inner parts only are preserved. 
This part of the hind wing is of a texture similar to that of the 
front wing. 

Length of front wing, 13 mm ; width, 4 ram . Type, No. 152. 

Lepium ? sp. 

A detached hind wing, contained in the collection, probably 
belongs with Lepium or with a closely related genus. The 
middle two-thirds of the wing is preserved, the extreme base 
as well as the apex being wanting. The sector divides tardily r 
resembling in this respect L. elongatum. Media is but two 
branched. Cubitus is a deeply impressed vein running from 
the base to the inner border. This vein gives off beyond the 
middle a thin, vaulted, forwardly directed, two branched 
division. The anal area is seen beneath and in front of the 
wing. It is thin, traversed by parallel veins, and lacks areola- 
tion. 

Width at the termination of the anal area, 4 mm . Width across 
the anal area probably not less than 4J to 5 mm . Length esti- 
mated, 11 to 12 mm . type, No. 319. 



E. H. Sellarcls — Types of Permian Insects. 157 

Atavci ovata gen. et sp. n. Text figure 3. 

The genus Atava is based upon a detached hind wing which 
differs so widely in venation as well as in size from other known 
types of this family as to represent without doubt a distinct 
genus. The radial sector is detached from the radius. It is 
three branched and fills the apex of the wing. Media is two 
branched. -Cubitus is a straight deeply impressed vein reaching 
from the base to the inner border. The anal area is wanting. 

Length of hind win^, 8 mm ; width at termination of the anal 
area, 3 mm . Type, ^0.^372. 

Lioraopteridm, family new. 

The family Liomopteridse includes a group of robust 
insects. The subcostal vein in this family is straight or nearly 
so, never arched, and gives off numerous oblique strong 
branches resembling in this respect many of the modern mantid 
species. The radius gives off oblique branches beyond the 
termination of the subcosta. The radial sector is several times 
branched. The media divides early. Cu 2 is simple. Cn, and 
the media are variable in their branching. 

The hind wing is shorter and broader than the front. The 
anal area is imperfectly known. It is, however, marked off by 
a deeplv impressed cubitus, and is doubtless expanded and 
folded. 

The legs are preserved, in part, on the type specimen and are 
seen to be relatively long and stout. The thorax is also partly 
preserved and is somewhat elongated. 

Liomopterum gen. n. 

The front wings of the species of this genus are elongate ; 
the costal border slightly arched. The subcosta lies in a shallow 
furrow and ends on the costa beyond the middle of the wing. 
The subcostal branches are numerous, oblique and mostly 
simple. The radius lies on a fold and reaches to the tip of the 
wing. Beyond the termination of the subcosta, numerous 
oblique simple veins pass from the radius to the border. The 
radial sector is given off at or near the termination of the basal 
third of the wing, and is three branched. The media divides 
early, one or both divisions being subdivided. The cubitus is 
uniformly arched at the base. Ciij is simple to three-branched ; 
Cu 2 is simple. 

The hind wing is thinner than the front. It is much 
expanded and is folded, the fold being marked off by a strong 
straight line. The radius of the hind wing is four branched, 
as is also the media. 



158 E. II SeUards — Types of Permian Insects. 

Liomopterum omatum sp. n. Text figures 5 and 6. 

M 2 of this species is simple. C\x 1 is three branched in the 
front wing. The hind wing is smaller and thinner than the 
front. The radial sector of the hind wing is four branched. 

Length of front wing, 14 mm ; width, 4 mm - Type, No. 5. 

Liomopterum extensum sp. n. 

The media of this species is four branched. Ciij is simple. 
Length of front wing, 14 mm ; width, 4f mm . Type, No. 972. 

Uorates elongatus gen. et sp. n. Text figure 2. 

A second and a somewhat larger genus is represented by the 
basal half of a front wing. The media divides in this genus 
much in front of the origin of the sector. Cubitus divides back 
of the bifurcation of media. The strong basal arch of the 
cubitus together with its late bifurcation leaves a ]arge area 
lying between the arch of the cubitus and the anal furrow. 
This area is traversed and strengthened by an accessory vein 
connecting the arch of cubitus with Cu 2 . The anal area is 
displaced and crowded partly across the wing. It is traversed 
as usual by two veins. 

Length of the wing (estimated), I7 mm ; width, 5 mn Y Type, 
No. 992. 

Trobnisidce, family new. 

The wings of the Probnisidse are of coriaceous texture, the 
membrane indistinctly wrinkled. Distantly placed cross veins 
are present. These are strong in the basal and inner part of 
the wing between the media, cubitus and anal veins, but are 
elsewhere weak and indistinct. The wings are long in propor- 
tion to their breadth. They are insufficiently braced, and were 
evidently imperfect organs of flight. The subcosta reaches to 
about the middle of the wing. The radius reaches to the apex. 
The sector arises early and is simple. Media divides near the 
base of the wing and is two to four branched. Cubitus is long, 
not infrequently reaching well toward the apical border. A 
variable number of inferior branches are given off by the 
cubitus. 

The hind wings are broader than the front. The sector is 
either detached or arises very early from the radius. The cubi- 
tus is a deeply impressed strong vein directed obliquely toward 
the inner border. A thin, forwardly directed, branched division 
is given off from the middle of the cubitus. The anal area of 
the hind wing is imperfectly known. The venation in this 
family presents many perplexing minor variations, and is prob- 
ably lacking in constancy. 



E. H. Sellards — Types of Permian Insects. 159 

Probnis speciosa gen. et sp. n. Text figures 10 and 11. 

The front and hind wings of the type specimen of this genus 
are preserved. The front wing is slender, being approximately 
three and a half times as long as broad. The membrane is 
coriaceous and wrinkled ; the veins lying in the membrane are 









Explanations of Text Figures 9-14. 

Fig. 9. Stinus brevi-cubitalis gen. et sp. n. Front wing. No. 459. 
Fig. 10-11. Probnis speciosa gen. et sp. n. Hind and front wings. No. 
143. 

Fig. 12. Stoichus elegans gen. et sp. n. Front wing. No. 974. 
Fig. 13. Espira obseura gen. et sp. n. Front wing. No. 101. 
Fig. 14. Stoichus arcuatus sp. n. Front wing. No. 973. 
All illustrations four times natural size. 



160 E. II. Sellards — Types of Permian Insects. 

thin and indistinct. The subcosta is arched at the base toward 
the costa and reaches to the middle of the wing. The radius 
reaches to the apex. The sector is simple. The media divides 
just in front of the origin of the sector, both divisions remain- 
ing simple. Cubitus is a very long vein reaching well along 
the apical border. It is thin and indistinct and gives off six 
slender branches. 

The hind wing is broader than the front, and is not more than 
three times as long as broad. The sector arises very early. 
The basal part of the cubitus is a heavy, strongly impressed 
vein, directed toward the inner border. From near the middle 
of the cubitus a thin forwardly directed branched division is 
given off which reaches to the apical border. The greater 
width of the hind wing results from the expansion of the inner 
border, and the greater area occupied by the cubitus. The 
anal area is unknown. The deeply impressed character of the 
cubitus leads one to expect a folded anal area. The membrane 
of the hind wing is of coriaceous wrinkled appearance similar 
to that of the front wings. 

Length of front wing about 13 mm ; width, 3 mm ; length of 
hind wing about 12 or 12i ram ; width, 4 mm . Type, No. 143. 

Prohiis coriacea sp. d. 

The cubitus in this species is scarcely so distinctly vaulted 
as in the case of the type species of the genus. Media divides 
opposite, rather than in front of, the origin of the sector. 

Length of front whig, 12 ram ; width, 3i mm . Type, No. 655. 

Espira obscura gen. et sp. n. Text figure 13. 

This genus is characterized by the very early origin of the 
sector ; by a three or four branched media ; and by a cubitus 
more vaulted and occupying a broader area, resulting in a pro- 
portionately broader wing than in the type genus. Espira resem- 
bles Prohiis in the late origin of the first cubital branch. It 
represents a type with broader and stronger wings, and with 
thickened membrane in which the veins can scarcely be traced. 

Length of wing about ll| mm ; width, 3-| mm . Type, No. 101. 

Stoichus elegans gen. et sp. n. Text figure 12. 

Media divides opposite the origin of the sector and is four 
branched. Cubitus is short and has beyond the first strong 
division but three branches. Cu x is vaulted at its separation 
from Cu 2 and is connected to the media by a strong cross vein. 
Cu 2 is much stronger than in the genera of this family already 
described. 

Length of front wing, ll ,nm ; width, 3f mm . Type, No. 974. 



E. H. Sellards — Types of Permian Insects. 161 

Stoichas arcuatus sp. n. Text figure 14. 

The first division of the media of this species is deeply 
forked, while the second division is simple. The cubitus is 
more strongly vaulted and is longer than is the cubitus of the 
type species. It has five slender branches, the third of which 
is forked. 

Length of wing, ll£ mm ; width, 3i mm . Type, No. 973. 

Stoichus minor sp. n. 

This species is very much smaller than the type species of 
the genus. Mediaj is simple ; M 2 is three branched. Cubitus 
gives off beyond the first strong branch, three inferior 
branches. 

Length of wing, 9 mm ; width, 3 mm . Type, No. 114. 

Stoichus tenuis sp. n. 

The wings of this species are thinner and the venation less 
distinct than in other species referred to this genus. The 
sector arises opposite the division of the media. Cubitus 
divides very early as in the type species. The offshoot from 
the cubitus is a thin vein reaching a little beyond the middle 
of the wing. Two inferior branches are given off, the first 
of which is forked. The first division of the media is simple, 
the second is apparentlv once forked. 

Length of front wing, 10 mm ; width, 3 mm . Type, JS T o. 1208. 

Stinus brevi-eubitalis gen. et sp. n. Text figure 9. 

The cubitus of this genus is short as compared with other 
genera of the family, being scarcely vaulted and reaching but 
slightly beyond the middle of the wing. Only two inferior 
branches are given off. The median area is correspondingly 
well developed. M 2 in the type species gives off three infe- 
rior branches which fill the apical border. M x is simple. The 
radial sector arises back of the division of the media, and is 
simple. 

Length of wing, 12| mra ; width, 4 mm . Type, No. 459. 

Lecopterum delicosum gen. et sp. n. 

This is a genus of small Probnisidse. The wing is slender 
and not so coriaceous as in other genera of the family. The 
veins in the apical part of the wing are thin and more or less 
wavy. Mj is simple. M 2 is three branched. Cubitus divides 
early. Beyond this first division it has but two inferior 
branches and is but slightly vaulted. 

Length of front wing, partly estimated, 9 mm ; width, 3 ,nm . 
Type, No. 824. 



162 E. 71. Sellards — Types of Permian Insects. 

Lemmatophoridce, family new. 

The family Lemmatophoridse includes small insects with 
four membranous wings as long as the abdomen. The pro- 
notum is bordered by a membranous expansion. The mesa- 
and meta-thoracic segments are strong. The sobcosta is 
simple, and terminates on the costal border near or beyond the 
middle line of the wing. Numerous oblique branches are 
given off from the radius beyond the termination of the sub- 
costa. The radial sector is simple to four branched. The 
media is weak and at the base lies very close to the radius ; 
media is two to four branched. The cubitus at the base has 
a strong upward curve toward the media, with which it is 
united by a few strong cross veins. Cu x is one to three 
branched. Cu 2 is simple. The anal area is marked off by a thin 
depressed line, and is traversed by one or two strong veins. 
Cross veins in the wing are comparatively strong although 
not numerous. 

The hind wings are broader and shorter than the front. The 
anal area of the hind wing is expanded and folded. 

JLemmatophora gen. n. 

This genus of small Lemmatophoridge has elongate mem- 
branous arched wings. The wing membrane is minutely scaly. 
Subcosta is not arched at the base and extends beyond the 
middle of the wing. Sector arises near the middle of the 
wing and is simple. M x is simple. M 2 is widely forked be- 
yond the middle of the wing, and is thin at its origin 
from M r Cu 3 is vaulted near its origin, and is two to three 
branched ; Cu 2 is simple. Two to four cross veins unite Rj 
with Rs. A strong cross vein unites Rs near its origin with 
Mj ; two to four other cross veins unite Rs and M r Ciij is 
united at the arch with M 2 by two or three strong cross veins. 
The anal area is traversed by two simple veins ; inner angle 
squarose. 

JLemmatophora typa sp. n. Text figure 24. 

Subcosta is united with the border by about nine cross veins. 
Two prominent cross veins occur between R x and Rs ; three 
to four cross veins between Rs and M,. M 2 is branched. Cu x 
is branched ; Cu 2 simple. The position of the main and cross 
veins of this species are very constant. Fourteen specimens 
are referred to the species. The branching of the main 
veins is without essential variation as is also the position 
and curvature of the veins. The cross veins of the wing are 
often obscure from lack of preservation, as is also the minutely 
scaly surface of the wing. 

Length of wing, 7 m,,i ; width, 2i mm . Type, No. 1162 
paratypes, Nos. 30, 32, 1266, 1376, 1377, 1379. 



E. II Sellards — Types of Permian Insects. 163 

Lemmatophora delicosa sp. n. 

This species resembles L. typa closely, both in size and in 
venation. M 2 is, however, three instead of two branched. 
Length of front wing, 7 mm ; width, 2i mm . Type, No. 1050. 

Lemmatophora anomala sp. n. 

One branch of the media of this species coalesces for a short 
distance with cnbitns. 

Length of wing, 7 to 8 mm ; width, 2^ ram . Type, No. 1089. 

lemmatophora hirsuta sp. n. 

This species is described from a specimen having the two 
pairs of wings and the body preserved. The prothorax is 
bordered by a membranous expansion l mm wide. The vena- 
tion of the wings is partly obscured as a result of the two 
pairs of wings lying together and partly over the abdomen. 
The costal border of the front wing is thickly set with short 
backwardly projecting spines or hairs. In other respects the 
venation so far as can be made out agrees with that of the type 
species of the genus Lemmatophora. 

Length of body from the prothorax to the end of the abdo- 
men, 10 mm ; length of prothorax about 2 or 2-| mm ; length of 
abdomen, 5^ mra ; length of front wings, 7 mm ; length of hind 
wings about 6i mm . Type, No. 1047. 

Lemmatophora elongata sp. n. 

This species has a more slender and a longer front wing than 
that of L. typa. 

Length of front wing, 9 mm ; width, 2| mm . Type, No. 149. 

Lisca minitta gen. et sp. n. Text figure 21. 

This is a genus of small Lemmatophoridge. The wing mem- 
brane is scaly and resembles in texture that of Lemmatophora. 
The radial sector originates not far from the middle of the 
wing. It is simple and is united for a short distance with M x . 
M ? is deeply forked. Cu x is forked. Cu 2 is simple. 

Length of front wing, 5i mra ; width, 2 mm . Type, No. 916. 

Artinska gen. n. 

The generic characters of this genus may be summarized as 
an arched subcosta, a bifurcate radius, and a minutely scaly 
wing membrane. 

The wing is of an elongate shape. The subcosta reaches 
somewhat beyond the middle of the wing. Oblique branches 
are given off from the subcosta and beyond the termination of 
the subcosta from the radius. The sector originates at the ter- 



164 K H. Sellards— Types of Permian Insects. 

Figs. 15-28. 






23 












E. H. Sellards — Types of Permian Insects. 165 

mination of the basal one-third or one-fourth of the wing. 
Media divides just in front of, opposite, or just back of the 
origin of the sector. Its subsequent divisions are variable with 
the different species. Cubitus divides in front of the division 
of the media. Its first division is variable ; the second is 
uniformly simple. The anal area is marked off by a thin line 
and is traversed by two veins. The cross veins are numerous 
although not always well preserved. The specific characters 
are taken for the most part from the median and cubital areas. 

Artinska clara sp. n. Text figure 25. 

The radial sector is deeply bifurcate. M t is simple. M 2 is 
deeply forked. The first division of the media occurs in front 
of the origin of the sector. Cu 3 is branched. Cu 2 is simple. 

Length of front wing, 10 mm ; width, 3f min . Type, No. 115. 

Artinska medialis sp. n. 

The media divides in front of the origin of the sector. The 
sector is shallow forked. M a and M 2 are both deeply forked. 
The first division of the cubitus is simple. 

Length of wing, 9 mm ; width, 3J mm . Type, No. 1381. 

Artinska gracilis sp. n. 

The sector is not deeply forked. M x is shallow forked ; M 2 
more deeply forked. Media divides opposite the origin of the 
sector. Cu, is three branched. 

Length about 9 mm ; width, 3i mm . Type, No. 1090. 

Artinska pecta sp. n. 

The sector is not deeply forked. Media divides in front* of 
the origin of the sector. M x and M 2 are deeply forked. 'CUj 
is three branched. 

Length of front wing, 10 mra ; width, 3-| min . Type, No. 1087 ; 
paratypes, Nos. 437 and 1068. 

Explanations of Text Figures 15-28. 

Fig. 15. Estadia elongata gen. et sp. n. Front wing. No. 1112. 

Fig. 16. Delopterum minutum gen. et sp. n. Front wing. No. 264. 

Fig. 17. Delopterum latum sp. n. Front wing. No. 94. 

Fig. 18. Hind wing, probably of one of the Lemmatophoridse. No. 1461. 

Fig. 19. Hind wing, probably of small species of Estadia, or related 
genus. No. 758. 

Fig. 20. Delopterum elongatum sp. n. Front wing. No. 61. 

Fig. 21. Lisca minuta gen. et sp. n. No. 916. 

Fig. 22. Prisca fragilis gen. et sp. n. No. 128. 

Fig. 23. Orta ovata gen. et sp. n. Front wing. No. 295. 

Fig. 24. Lemmatophora typa gen. et sp. n. Front wing. No. 1162. 

Fig. 25. Artinska clara gen. et sp. n. Front wing. No 115. 

Fig. 26. Leeorium elongatum gen. et sp. n. No. 524. 

Fig. 27. Stemma elegans gen. et sp. n. Front wing. No. 31. 

Fig. 28. Lectrum anomalum. Front wing. No. 173. 

Figures 15, 18, 19, and 26 enlarged five times ; all other illustrations four 
times natural size. 

Am. Jour. Sci. — Fourth Series, Vol. XXYII, No. 158. — February, 1909. 

12 



166 E\ II Sellards — Types of Permian Insects. 

Artinska major sp. n. 

The sector is deeply forked. Media divides in front of the 
origin of the sector. M x is simple ; M 2 is deeply forked. Cu r 
is three branched. 

Length of front wing, ll mm ; width, 4 mm . Type, No. 1375. 

Artinska extensa sp. n. 

The sector is shallow forked. Media divides much in front 
of the origin of the sector. M t is three branched ; M 2 is 
simple. Cuj is three branched. 

Length of front wing, 10 mm ; width, 4 mm . Type, No. 28. 

JEstadia gen. n. 

The generic characters may be summarized as an arched 
subcosta, and a radius which is bifurcate and is fused for a 
short distance with the first division of the media. 

The front wings are elongate. The subcosta reaches two- 
thirds of the length of the wing, and is strongly arched at the 
base. The sector arises 3 or 4 mm from the base and is fused 
for a short distance with M r Media is divided in front of the 
origin of the sector. Its subsequent divisions are variable. 
Cubitus 2 is simple. Cu x is bifurcate. The anal area is of the 
usual squarose form, and is traversed by two veins. The 
specific characters are taken, as in the other genera, from the 
disposition of the veins of the median and cubital areas. 

JEstadia elongata sp. n. Text figure 15. 

The radial sector is fused with M a for a distance of one milli- 
meter. M, is shallow forked ; M 2 is deeply forked. Cu, is 
forked. 

Length of front wing, 10i mm ; width, 3i ram . Type, No. 1112. 

JEstadia arcuata sp. n. 

The radial sector is fused with M x a distance of one-half 
millimeter. M> is shallow forked while M 2 is deeply forked. 
Cu x is two branched. 

Length of wing, 12 mm ; width, 4J mm . Type, No. 412; para- 
type, No. 358. 

JEstadia tenuis sp. n. 

The sector is fused with M x a distance of not more than one- 
half millimeter. M x is simple; M 2 is deeply forked. Cu x is 
two branched. 

Length of front wing, 9 mm ; width, 3i mm . Type, No. 1382; 
paratype No. 1274. 



E. H. Sellards — Types of Permian Insects. 167 

Lectrum anomalum gen. et sp. n. Text figure 28. 

The costal border of the front wing is gently arched. The 
snbcosta is but slightly arched at the base. The radial sector 
arises in front of the middle line of the wing. The veins of 
the wing are thin and not deeply impressed. Cu : is branched. 
Cu 2 is forked, differing in this respect from all other genera of 
the family. 

Length of front wing, estimated, 10 mm ; width, 3 mm . Type, 
No. 173. 

Prosaites compactus gen. et sp. n. 

The front wings of species of this genus are compact, the 
cross veins numerous. The radial sector is bifurcate. M, is 
simple ; M 2 is branched. Cu x and Cu 2 are both simple. The 
sector in the type species is united with media for a short dis- 
tance only. 

Length of front wing, 8J- mra ; width, 3 mm . Type, JSTo. 628. 

Prosaites secundus sp. n. 

The radial sector of this species is united with M x for a dis- 
tance of about one millimeter. M 2 is deeply forked. 

Length of front wing, 8i IQm : width, 3 mm . Type, No. 211. 

Prisca frag ills gen. et sp. n. Text figure 22. 

The front wing is elongate. The wing membrane is thin 
and the veins indistinct. The sector is simple and is united for 
a short distance with M r M x is simple. M 2 is bifurcate. Cu, 
is deeply forked. Cu 2 is simple. 

Length of front wing about 9 mm ; width, 3 mra . Type, No. 128. 

Lecorium elongatum gen. et sp. n. Text figure 26. 

The front wings are slender. The sector is forked at the tip. 
M x and M 2 are forked. Cu, is forked. Cross veins are present. 
Length of front wing, 8 mm ; width, 2| mm . Type, No. 521. 

Ortadce, family new. 

This is a family of small insects. The front wings are elon- 
gate, broadest beyond the middle. The subcosta reaches be- 
yond the middle line of the wing. The radial sector is usually 
two branched. The special peculiarity of this family is the 
union of media and cubitus at the base of the wing. The anal 
area is marked off by a thin line. Cross veins occur sparingly. 

The body is small, and is shorter than the wings. 



168 E. II SeUards — Types of Permian Insects. 

Orta ovata gen. et sp. n. Text figure 23. 

The radial sector is bifurcate. Media and cubitus are fused 
at the base for a distance of one to two millimeters. M x is 
simple ; M 2 is branched. Cubitus is divided at the point of 
separation from media. Cu x is deeply forked. Cu 2 is simple. 

Length of front wing, 7 mm ; width, 3 mra . Type, No. 295; 
paratype, No. 190. 

Stemma elegans gen. et sp. n. Text figure 27. 

The front wings are long and very slender. The sector is 
simple. Cubitus and media are united for some distance at 
the base. Cu 2 arises as a branch vein from the united main 
veins. M x is simple ; M 2 is three branched. Cu, is bifurcate 
at the tip ; Cu 2 is simple. 

Length of front wing, 10 mm ; width, 2i mm . Type, No. 31. 

Stemma extensa sp. n. 

The front wing is somewhat shorter than that of the type 
species. Cu a is simple. 

Length of front wing, 8 mm ; width, 2i mm . Type, No. 1272. 

Delopteridas, family new. 

This is a family of small insects. The front wings are nar- 
rowly elongate. The subcosta is short, rarely reaching beyond 
the middle of the wing. The radial sector is one to three 
branched. The cubitus, instead of dividing early into two 
main branches, continues simple until near the inner border, 
there giving off one or two thin, simple inferior veins. The 
anal area is marked off by a thin depressed line. Cross veins 
are not numerous in any part of the wing. 

The hind wings are of approximately the same length as the 
front. They are probably expanded and plicated as in other 
families of this group. 

The body is slender. The abdomen is much shorter than 
the wings. 

Delopterum minutum gen. et sp. n. Text figure 16. 

These are small insects. The subcosta is short. Radial sec- 
tor is three branched. Media is two branched. Cubitus is 
apparently simple. 

Length of wing, 4^ mm ; width, l£ mm . Type, No. 264. 

Delopterum elongatum sp. n. Text figure 20. 

This is a much larger species, and may possibly be found to 
be generically separable from Delopterum minutum. Cubitus 
in this species is branched. 

Length of front wing, f mm ; width, 2 ram . Type, No. 61. 



E. H. Sellards — Types of Permian Insects. 169 

Delopterum latum sp. n. Text figure 17. 

This species is broader and slightly larger than the type 
species. The sector arises very early. The sector is three 
branched. A fourth branch is given oif but is lost in the 
wing membrane. Cubitus is branched. 

Length of front wing, 5 mm ; width, If mm . Type, No. 94. 

Urba punctata gen. et sp. n. 

This genus is characterized by a slender elongate anal area 
traversed by three or four veins. The wing membrane has a 
punctate appearance, due probably to the presence of short 
spines bordering the veins. 

Length of front wing, 9 mm ; width, 3 ram . Type, No. HIT. 

Correlation of the Insect-bearing Horizon. 

The locality from which the insects described in this paper 
were obtained is three and one-half miles southeast of Banner 
City in Dickinson County, Kansas. The fossiliferous horizon 
occurs close to the top of the Paleozoic section of this part of 
the state and near the line of contact with the overlying 
Cretaceous. Fossil plants are associated with the insects. The 
matrix holding the fossils varies from an impure fine-grained 
laminated limestone to a hard concretionary limestone. Most 
of the insects were obtained from the laminated rock, while 
the plants come largely from the concretionary limestone. 
When the fossils were first discovered they were regarded as 
probably occurring in the Marion formation. 

As a result of stratigraphic studies made for the Kansas 
State Geological Survey during the past summer, J. W. Beede 
states that he has shown the plant and insect horizon to occur 
within the Wellington shales lying next above the Marion, 
and immediately under the Cretaceous. * 

Aside from doubtful forms, the paper contains descriptions 
of sixty species and thirty-five genera, all of which are new. 
To complete the faunal list, the cockroaches, not included in this 
paper, but described elsewhere, f should be added to this num- 
ber. Of this family ten species, all new, have been recognized. 
They are referred to two genera, one of which is new. Of 
groups larger than genera the following have been recognized 
in the Wellington shales : Odonata, Plecoptera, Megasecoptera, 
Oryctoblattinidse, Protorthoptera, and Paleoblattidse. 

The order Odonata is represented in the Wellington horizon 
by at least one form constituting a new genus and species. The 
Odonate phylum is known as early as the Coal Measures, being 

* Letter of Sept. 16, 1908. 

f Paper prepared for the Kansas State Geological Survey. 



170 E. H. Sellards — Types of Permian Insects. 

somewhat abundant and including forms of an unusually large 
size in the Commentry Coal Measures of France. Handlirsch 
regards the Coal Measure types of Protodonates as constituting 
an ordinal group of equal rank with modern Odonates. With 
this classification I have been unable to agree, as the specimens, 
from the Wellington shales at least, possess the essential 
ordinal characters of true Odonata. 

The order Plecoptera, or ephemerids, is somewhat abundant 
in the Wellington shales. In Part II of this paper I have 
described ten genera and thirteen species constituting a new 
family of this order. Insects which appear to be prototypes 
of the ephemerids exist in some abundance in the Coal Meas- 
ures. Handlirsch * has recognized ephemerids as occurring 
sparingly in the Permian of Russia. With this exception, 
true ephemerids have not previously been identified from 
Paleozoic deposits. The relative abundance of this group of 
insects in the Wellington shales affords an exceptionally strong 
argument for the Permian age of that formation. That the 
members of this group, as it occurs in the Wellington, are 
provided with two pairs of fully developed wings and are 
otherwise far more primitive than modern ephemerids, by no 
means weakens the argument, since these are precisely the 
characters to be expected in early members of the order. 

The order Megasecoptera is sparingly represented in the 
Wellington shales, a single specimen having been obtained. 
This order was described from Coal Measure specimens, from 
which deposit alone it has been known heretofore. The 
continuance of the order into the Permian, however, is not 
unexpected. Aside from cockroaches, relatively few genera 
of insects have been described from the Permian. This fact 
is sufficient to account for the previous lack of knowledge 
regarding the continuance of the Megasecoptera into the Per- 
mian. 

The family Oryctoblattinidse was established upon Coal 
Measure and Permian forms, seven Coal Measure and two 
Permian genera having been referred to the group. From the 
Wellington shales two genera of this family, both new, are 
described in this paper. 

Protorthoptera is the predominant order of insects in the 
Wellington shales. This order is recognized as common to both 
Coal Measure and Permian deposits. Six families of the order 
are described in this paper, all of which are new. The forms 
making up these six families constitute twenty genera and 
forty-three species. 

The cockroaches of the Kansas Permian have been described 
in a paper now being published in the reports of the Kansas 
* See Pt. II, p. 345. 



E. B. Sellards — Types of Permian Insects. 171 

State Geological Survey. A detailed account of the group is 
therefore omitted from this paper. It has usually been observed, 
in collecting from Paleozoic localities, that cockroaches exceed in 
number of individuals all other insects combined. In the 
Wellington shales the cockroaches are much in the minority. 
A collection of something over two thousand insect specimens 
was found to contain only about seventy cockroaches. From 
these, two genera and ten species were identified. Of the two 
genera, one is the well known Coal Measure and Permian genus 
Etoblattina. The second genus is new. The ten species ob- 
tained are new. The rarity of cockroaches in the Wellington 
is in marked contrast to their relative abundance in most Coal 
Measure and early Permian localities. 

Fossil insects have been obtained from the Birmingham 
shales near Steubenville and Kichmond, Ohio, in the Cone- 
maugh series just above the Ames or Crinoidal Limestone. 
Recently reptilian remains heretofore supposed to be Permian 
have been found near Pitcairn, Pennsylvania, in the red clay 
below the Ames limestone.* The presence of the reptilian 
remains has given rise to a question as to the age of the Cone- 
maugh series. In the collection of insects from that locality, 
Scu elder recognized twenty-two species referable to three genera. 
All of these are cockroaches, other families not having been 
found at this locality. Of these twenty-two species, seventeen 
were referred by Scudder to the genus Etoblattina, three to 
Gerabldttina, and two to Povoblattina. Of these genera 
Etoblattina alone is recognized in the Wellington shales, and 
as already remarked, the cockroach family is in the minority at 
that locality. E"o one of the twenty-two species of the Kich- 
mond locality has been recognized in the Wellington. On the 
other hand, two of Scudder's species, Etoblattina maladicta 
and E.benedicta (regarded by the writer as a single species 
and referred to Sjpiloblattina), have been obtained from the 
Leroy ( Coal Measure ) shales of Kansas.f It should be added, 
however, that Handlirsch^: does not agree with the writer either 
in uniting these two species, or in identifying them with the 
specimens from the Leroy shales. Aside from the question of 
discrimination of species, about which there may be differences 
of opinion, the essential fact remains that a closely similar 
type of wing development is seen in species from the two 
localities. This type of wing venation is referred to by Scud- 
der as the " remarkable openness of the neuration in the middle 
of the tegmina.§ Handlirsch assigns to the cockroaches of this 

*P. E. Kaymond, Science, xxvi, p. 835, 1907. 
fThis Journal, vol. xviii, p. 214-216, 1904. 
JDie Fossilen Insekten, p. 240, 1906. 
§ Bulletin U. S. Geol Survey No. 124, p. 12, 1895. 



172 E. H. Sellards — Types of Permian Insects. 

type, family rank, the Spiloblattinidse. The family as thus de- 
limited predominates at the Steubenville and Richmond local- 
ities, and is known as high as the Cassville plant shales at the 
base of the Dunkard series of West Virginia. It is abundant in 
the Leroy shales of the Kansas Coal Measures, but has not 
been recognized in the Wellington shales at the top of the Kan- 
sas section. In Europe this group is reported from both 
Upper Carboniferous and Lower Permian. The type genus, 
/Spiloblattina, was described from Fairplay, Colorado. 

The Fairplay locality was placed by Scudder in the Trias. 
Lesquereux maintained, however, that from the plant evidence 
the formation could not be later than Permian. With regard 
to the plant material, Mr. David White states :* " The plant 
and insect beds at Fairplay, referred by Doctor Scudder to the 
Trias, and by Lesquereux to the Permian, can, on the evidence 
of the plants, not be regarded as later than Permian, if indeed 
they are above the highest Coal Measures." The insect remains 
as now interpreted are not in conflict with the plant evidence, 
and certainly do not require the reference of the formation to 
the Triassic. 

Of the two remaining genera occurring at the Ohio locality 
Etoblattina is a Coal Measure-Permian genus. Poroblattina 
is found at Fairplay, Colorado, and in the upper Carboniferous 
and Lower Permian of Europe. The insect remains thus far 
obtained do not therefore permit a close correlation of the 
Birmingham shales with the Kansas section. It seems prob- 
able, however, that that formation is of somewhat later age 
than the Leroy shales of the Kansas Coal Measures. 

A number of insects have been obtained, principally by 
Mr. P. D. LaCoe from the Cassville plant shales at Cassville, 
West Virginia. This horizon lies at the base of the Dunkard 
series variously regarded as Permian or as Permo-Carbon- 
iferous, and occurs, according to I. C. White, some six hun- 
dred feet in the stratigraphic column above the Birmingham 
shales. In the collections from the Cassville locality, Scudder 
recognized fifty-six species referable to five genera all of which 
are cockroaches. I have been unable to recognize in the Wel- 
lington shales the presence of any one of the fifty-six species 
occurring at the Cassville locality, and only one genus Etoblat- 
tina is common to the two horizons. The predominance of 
the cockroach fauna together with the absence of such advanced 
types as true ephemerids, leads to the view that the Cassville 
locality, although possibly Permian, is much older than the 
Wellington shales of the Kansas section. 

Among the few insects obtained from the Permian for- 
mation of Russia, Handlirsch recognizes, as previously stated, 
*TJ. S. Nat. Mus., vol. xxix, pp. 687, 1906. 



E. H. Sellards — Types of Permian Insects. 173 

the occurrence of true ephemerids. The Russian deposits have 
also yielded forms regarded as representing Paleohemiptera 
and Mantoidse.* These last two groups have not been recog- 
nized in the Kansas Permian. The presence of the ephemerids, 
however, forms a strong tie in common between the insects of 
the Russian and the Kansas Permian. 

The insects of the Wellington are on the average of small 
size as compared with Coal Measure insects. This is particu- 
larly noticeable among the cockroaches, all of which are small. 
This dwarfing of the fauna is of interest as probably indicat- 
ing unfavorable climatic conditions. 

The Geological Relations of the Associated Plants. 

Plants, as previously stated, are associated with the insects 
at the Banner City locality. A paper describing the plants 
from this horizon is being published by the Kansas Geo- 
logical Survey. In the writer's opinion, the plant fossils 
indicate unequivocally the Permian age of the formation from 
which they come, The evidence as to the age of the Wel- 
lington shales, derived from the flora, is thus summarized in 
the report referred to ; "More than two-thirds of the Welling- 
ton species are either identical with or most closely related to 
species or genera characteristic of the European Permian. 
The points which seem to have the most importance as 
bearing on correlation of the Wellington are the following : 

(1) The complete absence from the Wellington of species in 
any way confined to or distinctive of the Coal Measures. 

(2) The comparatively small number of species originating as 
early as Upper Coal Measure time. (3) The presence of a 
few species common to and characteristic of the Permian of 
Europe. (4) The close relation of the new forms to species 
characteristic of the European Permian. (5) The distinctly 
Permian facies of the flora as a whole and its marked advance 
over the flora of the Upper Coal Measures. 

The advance in the flora consists in the number of species 
and in the abundance of individuals of callipterid and tseniop- 
terid ferns, and of the new fern genus, Glenopteris, which 
appears to be related, on the one hand, to callipterid ferns of 
Permian types, and, on the other, to the Triassic genera 
Cycadopteris and Lomatojpteris. 

The evidence derived from the fossil plants as a whole 
seems to assure the reference of the Wellington to the true 
Permian in the European sense." 

This conclusion drawn from the plant fossils is now fully 
confirmed by the evidence derived from the insects. 

* Die Fossilen Insekten, 348, 1906. 



174: G. Edgar — lodometric Estimation of Vanadic Acid. 

Art. X. — The lodometric Estimation of Vanadic Acid, 
Chromic Acid and Iron in the Presence of One Another / 
by Graham Edgar. 

[Contributions from the Kent Chemical Laboratory of Yale Univ. — cxciv. ] 

In a previous paper from this laboratory* it has been shown 
that vanadic and chromic acids may be estimated in the presence 
of one another by a process based upon the differential reducing 
action of hydrobromic and hydriodic acids. The present paper 
will show that the use of processes of differential reduction 
may be extended to the estimation of three constituents ; in the 
present case, vanadic acid, chromic acid and iron. 

A preliminary investigation was made in order to determine 
whether vanadic acid and iron might be estimated iodometrically 
by distillation, first with hydrobromic acid and last with hydri- 
odic acid, the liberated halogen being absorbed in potassium 
iodide and determined after each distillation. In the first dis- 
tillation the vanadic acid should be reduced to the tetroxide, 
according to the equation, 

V 2 5 + 2HBr = Vfi 4 + H 2 + Br 2 . 
In the second distillation, the vanadium tetroxide should be 
reduced to trioxide and the ferric salt to ferrous salt, according 
to the equation, 

V 2 4 + Fe 2 3 + 4HI = V 2 3 + 2FeO + 2H 2 + 2l 2 . 
It is evident that both constituents may be calculated from 
the amount of halogen liberated in the two reductions. 
Experiments upon solutions of sodium vanadate and ferric 
chloride, carried out in the manner to be described later, gave 
the results shown in Table (I). 

If this process be carried out in the presence of chromic 
acid the reduction by hydrobromic acid should proceed accord- 
ing to the equation, 

V 2 5 + 2Cr0 3 + 8HBr = V 2 4 + O 2 3 -I- 4H 2 + 4Br 2 , 
while the reduction by hydriodic acid should proceed as before, 
according to the equation, 

V 2 4 + Fe 2 3 + 4HI = V 2 3 + 2FeO + 2H 2 +2l 2 . 

If the halogen liberated in the two reductions be separately 
determined, we have two equations, the first giving the sum of 
the vanadic and chromic acids, and the second giving the sum 
of the vanadic acid and the iron ; the halogen liberated by the 
vanadium being the same in each case. If then either the 
vanadium, iron or chromium be estimated separately, we obtain 
a third equation, from which, with the aid of the first two, all 
three constituents may be calculated. 

In this investigation an attempt was made to estimate the 
vanadium in a separate portion of the solution by reduction 

* Edgar, this Journal, xxvi, p. 333, 1908. 



G. Edgar — Iodometric Estimation of Yanadic Acid. 175 

with sulphur dioxide, boiling out the excess of reagent, adding 
barium chloride to precipitate the sulphuric acid formed, and 
then subjecting the solution to distillation with hydriodic 
acid. The chromic acid and the iron being already reduced 
by the sulphur dioxide, it is evident that the iodine liberated 
in the distillation should correspond to a reduction of vanadium- 
tetroxide to trioxide. This process, however, yielded faulty 
results, due, as was ascertained, to the formation of some dithi- 
onic acid during the treatment with sulphur dioxide. 

Another attempt was made to treat the residue after distil- 
lation with hydriodic acid with an excess of standard iodine 
solution and make alkaline with sodium bicarbonate, the excess 
of iodine to be afterwards determined with standard arsenious 
acid. In this process the vanadium trioxide should be oxidized 
to pentoxide and the ferrous salt to ferric salt, thus yielding 
the necessary data for the third equation. This method 
was found to be unavailable, owing to the fact that in the 
presence of the bulky precipitate of chromic and ferric hydrox- 
ides the vanadium is reoxidized very slowly by iodine. 

The method which finally proved successful was the esti- 
mation of the chromic acid in a separate portion by a modifi- 
cation of Browning's method of reduction with standard arseni- 
ous acid.* 

The experiments to be described were made upon mixtures 
in varying proportions of solutions of sodium vanadate, potas- 
sium bichromate and ferric chloride ; the entire process in 
detail being the following : 

(I). The solution, of about fifty cubic centimeters volume, 
was divided into two equal parts, one of which was placed in a 
distillation flask. To this, one to two grams of potassium 
bromide were added, together with twenty-five cubic centi- 
meters of concentrated hydrochloric acid, and the mixture was 
distilled until a volume of about twenty -five cubic centimeters 
was reached, the reduction having always been found to be 
complete in that time. The bromine liberated was absorbed in 
alkaline potassium iodide, and, after cooling and acidifying, the 
iodine liberated was estimated by titration with approximately 
tenth normal sodium thiosulphate. The results are given in 
(I), Table (II). 

(II). The absorption apparatus was replaced, and after the 
addition to the solution in the distillation flask of one gram of 
potassium iodide, ten cubic centimeters of concentrated hydro- 
chloric acid and three to five cubic centimeters of syrupy phos- 
phoric acid, distillation was again carried on until a volume 
of ten cubic centimeters had been reached ; the iodine thus 
liberated being estimated as before. The results are given 
under (II), Table (II). 

* This Journal, vol. i, p. 35, 1896. 



176 G. Edgar— Iodometric Estimation of Vanadic Acid. 

(III). In the second half of the original solution the chromic 
acid was estimated by adding sulphuric acid to slight acidity, 
three cubic centimeters of syrupy phosphoric acid and an 
amount of standard arsenious acid in excess of that required to 
effect the reduction of the chromic acid. After standing for 
fifteen to twenty minutes the solution was made alkaline with 
sodium bicarbonate, and an excess of standard iodine solution 
was added. This was allowed to stand in a stoppered flask for 
from one half hour to an hour, the excess of iodine being then 
removed with arsenious acid and the solution titrated to color 
with iodine after the addition of starch. 

The use of phosphoric acid causes the iron to be precipitated 
as phosphate, and thus the difficulty mentioned by Browning* 
in observing the end-point, due to the reddish-brown ferric 
hydroxide, is in large measure obviated, the blue of the starch 
iodide being quite clear against the pale green chromic hydrox- 
ide. The results are given in (III), Table (II). 

The first step in the calculation is the reduction to terms of 
tenth normal solution of the figures under (I), (II) and (III). 
It is evident that the subtraction of (III) from (I) gives the 
number of cubic centimeters corresponding to reduction of the 
vanadium pentoxide to tetroxide, while the subtraction of this 
result from (II) gives the number equivalent to the reduction 
of the ferric salt. By multiplying these figures by the solutions' 
equivalents of vanadic acid, chromic acid and ferric oxide, the 
necessary data are obtained. An example will make this 
clearer : 

cm 3 N/10 factor cm 3 

(No. 1) (I) 31-15 X l'lOO = 34-26 (N/10) =-== V 2 5 + Cr0 3 

(II) 23-4 X MOO = 25-74 (N/10) =^= V 2 6 + Fe 2 3 

III 30-00-8-74 (N/10) = 21-26 (N/l 0) -*c= OrO, 
(I)-(III) (34-26 - 21-26) = 13*00 cm 3 =o= V 2 5 

(II)-13-00 (25-74 - 13-00) = 12-74 cm 3 =-j= Fe 2 3 

21-26 X 0-003334 (factor for Cr0 3 ) = 0*0709 grm. Cr0 3 found. 
13-00 X 0-00912 (factor for V 2 5 ) = 0*1185 grm. V 9 B found. 
12*74 X 0*00799 (factor for Fe 3 O s ) = 0-1018 grm. ¥e.O s found. 

The distillation flask used in this work will be briefly de- 
scribed. It consisted of a 100 cm3 pipette modified as shown in 
fig. (I), the inlet tube being bent upwards and having a separ- 
atory funnel sealed to its end, while 'the outlet tube was bent 
upwards and then down to enter the absorption flask ; a small 
bulb being blown in it to prevent mechanical loss during distil- 
lation. In carrying out the process a slow current of hydro- 
gen from a Kipp generator was kept up through the apparatus, 
and this entering near the bottom of the flask enabled distil- 
lation to be carried to very small volume without danger of 

* Loc. cit. 



G. Edgar — Iodometric Estimation of Yanadic Acid. 177 

" bumping." The flask is similar in design to that used by 
Professor Gooch for the determination of boric acid and has 
been convenient and serviceable not only for this but for 
other processes involving distillation ; it being very simple 
and doing away with the necessity of ground glass connections. 




Summary. 

It has been shown that vanadic acid, chromic acid and iron 
may be accurately estimated in the presence of one another by 
an iodometric process based upon the differential reducing 
action of hydrobromic and hydriodic acids, upon the substances 
in question ; one constituent, the chromium, being estimated 
in a separate portion by reduction with arsenious acid. 

Table (I). 















(I) 


(II) 














N/lOx 


N/lOx 


Taken 


Found 


Error 


Taken 


Found 


Error 


•9968 


•9968 


v 2 o 5 


V 2 5 


v 2 o 5 


Fe 2 3 


Fe 2 3 


Fe 2 3 


Na 2 S 2 3 


Na 2 S 2 3 


grm. 


grm. 


grm. 


grm. 


grm. 


grm. 


cm 3 . 


cm 3 . 


0-1532 


0-1532 


±0-0000 


0-1515 


0-1515 


±0-0000 


16*85 


35-88 


0*1532 


0-1532 


±0-0000 


0-1515 


0-1518 


+ 0-0003 


16-85 


35-91 


0-1532 


0-1531 


—o-oooi 


0-1515 


0-1518 


+ 0-0003 


16-84 


35-90 


0-1532 


0-1531 


— 0-0001 


01515 


0-1516 


+ o-oooi 


16*84 


35*88 


0-1532 


0-1535 


+ 0*0004 


o-ioio 


o-iooi 


— 0-0009 


16-89 


29*46 


0-1532 


0-1533 


+ 0-0001 


0-1010 


0-1007 


— 0-0003 


16-86 


29-50 


0-0511 


0-0509 


— 0-0002 


0-1515 


0-1516 


+ 0-0001 


5-60 


24-64 


00511 


0-0511 


±0-0000 


0-1515 


0-1515 


±0-0000 


5-62 


24-65 


0-2042 


0-2045 


+ 0-0003 


0-0505 


00502 


— 0-0003 


22-50 


28-80 


0-2042 


0-2042 


±0-0000 


0-0505 


0-05,05 


+ 0-0000 


22-47 


28-82 



178 G. Edgar — lodometric Estimation of Vanadio Acid. 



r 


o 


d 1 




o 


© 


o - 


o 


o 


o 


o 


© 


© 


© 


© 


© 


i 






o 


b 


o 


o 


o 


b 


b 


b 


© 


b 


© 


© 


i 
i 

I 


fc 


<^ 


CO 


CO 


CO 


T* 


TtH 


TJH 


<M 


CM 


»o 


CM 


CM 


CO 


o 






^H 


o 


K> 


© 


lO 


o 


id 


O 


id 


© 


lO 


© 


l 


£ 


res 

' o 

M 


6o 


oc 


00 


00 


00 


r— 1 
(30 


CO 

6s 


OS 




os 


CO 

os 


00 


X 


o 


6 




© 


o 


CM 


»o 


o 


o 


as 


lO 


io 


© 


© 


© 


fc 


o 


CO 
OS 




T^ 1 

CO 


1? 

CO 


CO 


cm 

OS 


<M 

© 




•1— t 

© 


CO 


»p 


© 


CM 

id 


CO 




CM 


CM 


CM 


CM 


CM 


rH 


CM 


CM 


r-H 


<* 


CO 


""* 


X 




6 




to 


CM 


co 


o 


lO 


o 


00 


o 


00 


CM 


lO 


»o 


oo 
1—1 v* 


CAD 

OS 




i— i 


i— 1 
i— 1 ' 


r-H 


00 

6 


© 


© 

4* 




r-H 


4h 


CM 

CO 


(M 

CO 


I— 1 

I— 1 


£ 




CO 


CO 


CO 


-* 


^ 


CO 


I— 1 


CM 


^H 


CO 


CO 


CO 










CO 


CM 


CO 


CM 


CO 


CM 


>— 1 


CO 


CO 




-* 


I— 1 




rH 


6 

c* 




o 


O 


o 


o 


o 


o 


o 


o 


© 


© 


© 


© 






o 


o 


o 


o 


o 


o 


o 


o 


© 


© 


© 


© 




d 


o 


o 


o 


o 


© 


o 


o 


© 


© 


© 


o 


© 






Mi 


o 


o 


o 


o 


o 


o 


o 


© 


© 


© 


© 


© 








1 


+ 


1 


+ 


+ 


1 


+ 


1 


+ 


1 


1 


+ 



cocooOeo^o5cooo^©J>-j— i 

^f-s.— Clr-( CO CO rH^^H CM CO nH-H 

g^g ©©©ioio©©©©»o©*o 

§©£7* >-H T-H ,_, r-H ,_, CM T-H r-H r- O © 

fe^ M o © © b © © © © © © © © 

i^^^T-H!^rHCMrHi--!r-l»-H© 
PJ /-? J <M CM <M CO CO CM Th CM <M CO CM r— 1 

S«S° o © 10 xo © © © © 10 © 10 



r^ 



© 



S r* M) • • • . . . . . . - 

H m ©©©©©©©o©©©© 



9 




© 




© 


r— i 


CM 


r-H 


r-H 


I— 1 


© 


CM 


r-H 






© 


© 


© 


© 


© 


© 


© 


© 


© 


© 


© 


© 




^ » . 


© 


© 


© 


© 


© 


© 


© 


© 


© 


© 


© 


© 




go g 


© 


© 


© 


© 


© 


© 


© 


© 


© 


© 


© 


© 




HO M 


© 


© 


© 


© 


© 


© 


© 


© 


© 


© 


© 


© 




-H 


+ 


+i 


+ 


1 


+ 


+ 


1 


+1 


1 


+ 


+ 




'S » 


OS 


© 


© 


iO 


CM 


IO 


IO 


CO 


r^ 


CM 


IO 


© 




© 


r-i 


c 


CO 


CO 


CO 


IO 


10 


r-H 


IO 


IO 


r-H 




30 a 


J^ 


J> 


i^ 


© 


© 


© 


CO 


CO 


"* 


CO 


CO 


-t^ 




o.hS 


© 


© 


© 


r-H 


r— 1 


r—i 


© 


© 


r-H 


© 


© 


© 




£0 bD 


© 


© 


© 


© 


© 


© 


© 


© 


© 


© 


© 


© 




5°^ 


os 


© 


© 


"# 


xH 


tH 


** 


T*i 


x^ 


■<* 


rjH 


© 




© 


© 


© 


CO 


CO 


CO 


10 


10 


1— 1 


iO 


10 


© 




x>- 


C^ 


i^ 


© 


© 


© 


CO 


CO 


Tfl 


CO 


CO 


i^ 





© 


© 

© 


© 
© 


© 


r-H 
© 


r— 1 
© 



© 


© 

© 


r-H 
© 


© 
© 



© 


© 
© 



CM ^ CO r— I©-— I I— I IO — IrHlOr-H 

©©©0©©©©© l( ^>©© 
r!^?--© ° O © © © © © © CO © © 

gogo © © © © © © O © © © © 

H +I+ + -H|| + |II! 

rrj iO© l- O^CO©©00©»0'— < CM 

00i^000000©©00©©©00 



SOSr-H r-H r-H r-H r-H »0 lO r-H IO CO CO 

o^bcr r t 1 r r 1 9 9 t 1 9 ^ ?" 

fn b©bbb©b©©o©© 

_, COCOCOCOCOr-^r-HCOr-H©COCO 

S^flOO 00 00 OO OC © © GO © CO CO 00 

^OS^ rHr ^_ Hr ^ i OiOr-H 1 OCOCOr^ 

*i>tjD ^r , T7 H r99T H 9 < ? ,< ^'^ 

^ bbbbbbbbb©©© 

d r-HCM"co-^xdcd-t--oo©©i—i(M 



Bradley — Composition of the Mineral Warwickite. 179 

Art. XI. — On the Analysis and Chemical Composition of 
the Mineral Warwickite; by W. M. Bradley. 

Historical. — The mineral warwickite was first described by 
Shepard* in 1838 and again more fully in 1839 ; in the latter 
article he describes his method of analysis. Shepard named 
the new species warwickite, after the original locality, War- 
wick, Orange Co., JN\ Y. The mineral was found in limited 
quantity as small, slender crystals, imbedded in a highly crys- 
talline white limestone. It had earlier been called hypersthene 
on account of the brilliant copper-red reflections afforded by 
its cleavage surfaces. At a second occurrence found by Young 
and Horton in the vicinity of the first, pieces about half an 
inch in diameter were obtained. These latter crystals lacked 
the copper-red luster characteristic of those from the first- 
mentioned locality, and were in a more or less decomposed 
condition. 

From a qualitative analysis Shepard concluded that war- 
wickite was a fluo-titanate of iron and manganese with a small 
percentage of yttrium. His results, however, were shown 
later (cf. Smith, noted below) to be erroneous and need not be 
discussed here. 

In 1846, Huntf published an article on a supposed new spe- 
cies from Warwick, ~N. Y., which he called enceladite. The 
material analyzed (I, below) was the impure altered warwickite 
examined by Shepard, as Hunt himself recognized later. His 
second analysis:}; (II, below) was made on a purer specimen but 
showed a loss of nearly 20 per cent, which he attributed to 
an accident. Hunt's analyses are as follows : 

I II 

Si0 2 18-50 

Ti0 2 28-20 31-50 

FeO 10-59 Fe 2 3 8-10 

MgO 22-20 43-50 

A1 2 3 ; 13-84 

CaO.._ 1-30 

H 2 7-35 Ignition 2-00 

101-98 

After the completion of Hunt's analyses several years 
elapsed before further work was done on this mineral. About 
1853, Brush and Smith, then engaged in the re-examination 

* This Journal (1), xxxiv, 313, 1838, xxxvi, 85, 1839. 

f This Journal (2), ii, 30, 1846. % Ibid., xi, 352, 1851. 



180 Bradley — Composition of the Mineral Warwickite. 

of American minerals, pointed out that warwickite possessed 
a peculiar composition altogether different from what had 
been supposed, and that the pure mineral had not yet been 
analyzed. When the mineral had been subjected by the 
gentlemen named to a careful qualitative analysis, it was 
found to their great surprise to contain a large amount of boric 
acid, hitherto overlooked, so that it was to be considered as a 
boro-titanate of magnesia and iron. The quantitative analysis 
undertaken by Smith was made with difficulty because of the 
small amount of material available and from the fact that 
minute crystals of spinel penetrated those of warwickite. 
Smith, however, was finally satisfied that the results obtained 
expressed the true composition. The specific gravity obtained 
(Brush) was 3*362, and Smith's analysis is as follows : 

Oxygen Eatio 

B 2 3 _-___ 27-80 19-06 9 

Ti0 2 23-82 10-37 5 

MgO 36-80 14-46 6 

FeO 7'02 2-10 1 

Si0 2 1-00 

A1 2 G 3 ____ 2-21 

98-65 

Smith regarded , the silica and alumina as impurities, the 
latter arising from the spinel that it had been impossible to 
separate ; this, with a little of the magnesia, he deducted in 
making out the oxygen ratio from which he derived the 
formula : 

6MgO . lFeO . 2Ti0 2 . 3B 2 3 . 

Method of Analysis. — The material used for the present 
analysis was obtained from the Brush collection and came from 
Amity, N. Y., where this mineral is found as a characteristic 
associate of the granite contacts of the region. The warwick- 
ite occurs in minute slender crystals showing the copper-red 
reflections of the cleavage surfaces which is so characteris- 
tic of the pure mineral. It is found in a coarsely crystalline 
white limestone, intimately associated with a greenish blue 
spinel, black spinel, magnetite, serpentine, chondrodite and 
occasional scales of graphite. The limestone rock containing 
the minute crystals of warwickite was crushed to small frag- 
ments and these small pieces, which contained some of the 
mineral, were carefully selected by means of a glass. This 
material was again crushed and prepared for treatment with 
heavy solutions. Potassium mercuric iodide solution, having 
when concentrated a specific gravity of 3*15, was first used to 
separate the greater part of the calcite and serpentine. The 



Bradley — Composition of the Mineral WarwicMte. 181 

final separation was made by means of barium mercuric iodide 
with a specific gravity of 3*55. 

Considerable difficulty was caused by the presence of a 
greenish blue glassy spinel which in the solution closely 
resembled the grains of warwickite, but by repeated treat- 
ment with the heavy solution the latter was obtained in a quite 
pure condition. The material was further purified by the action 
of an electro-magnet which helped to remove some of the 
remaining foreign material. Finally by means of a very power- 
ful glass the few remaining grains of the associated minerals 
were as far as possible removed. The final sample obtained 
amounted to a little over two grams and was quite uniform in 
character. 

The specific gravity, determined by means of the barium 
mercuric iodide solution and a Westphal balance, was found 
to be 3 - 342 ; this is practically the same as that given by 
Brush, viz. : 3*351 for small fragments. 

Owing to the limited amount of material available for 
analysis it was desirable to determine the main constituents, 
B 2 3 , Ti0 2 , MgO and total iron in one portion. After 
repeated fusions of the mineral with sodium carbonate the 
resulting cake was soaked out and the liquid decanted through 
a filter, the residue being thoroughly boiled with 25 cc of 
sodium carbonate solution and transferred to the filter and 
finally washed with dilute sodium carbonate solution. The 
filtrate containing the boron was transferred to a distilling 
bulb and the determination of boron made by distilling with 
methyl alcohol, the distillate being collected in ammonium 
hydroxide and finally evaporated over calcium oxide. The 
residue left in the bulb after distillation contained a trace of 
titanium which was recovered and added to the main solution 
previous to the precipitation of the titanium. The residue 
from the sodium carbonate fusion was brought into solution 
by prolonged fusion with acid potassium sulphate, and the 
resulting cake dissolved in cold water to which had been 
added strong S0 2 water. The solution was then largely 
diluted and rather strongly acidified with acetic acid ; ' the 
titanium precipitation being made in the presence of sodium 
acetate and brought about by boiling the solution from three 
to five minutes, strong S0 2 water being added before the boil- 
ing point was reached. The precipitate was then filtered and 
washed with dilute acetic acid and finally weighed as Ti0 2 . 
Some of the details of the above briefly outlined method are 
those recommended by Warren. * 

The filtrate from the precipitation of titanium was concen- 
trated and a very small precipitate was collected and added to 

* This Journal (4), xxv, 23, 1908. 

Am. Jour. Sci.— Fourth Series, Vol. XXVII, No. 158.— February, 1909. 
13 



182 Bradley — Composition of the Mineral Warwickite. 

the main precipitate before igniting. A trace of iron was also 
precipitated at this point, and this together with a mere trace 
retained by the main titanium precipitate was recovered by 
fusing the Ti0 2 with acid potassium sulphate and a volumetric 
determination for iron was made in the usual way with 
KMn0 4 . The filtrate from the titanium precipitation con- 
taining the iron in the ferrous state was treated with nitric 
acid to oxidize the iron, and hydrochloric acid added to form 
enough ammonium chloride to keep the magnesium in solution, 
when ammonium hydroxide was added to precipitate the iron, 
etc. Double precipitations of the hydroxides were made and 
the weight of the mixed oxides obtained. The oxides were 
then fused with acid potassium sulphate, and the total iron 
determined as usual by titration with KMn0 4 . Traces of 
titanium retained by the precipitate of ferric hydroxide were 
determined where present by the colorimetric method and 
corrections for both iron and titanium were made. The 
amount of the alumina present was as usual arrived at by 
difference. The filtrate from the ammonium hydroxide pre- 
cipitation served for the determination of magnesium, which 
was precipitated as ammonium magnesium phosphate. This 
was then dissolved, reprecipitated and filtered on a Gooch 
crucible and finally weighed as magnesium pyrophosphate. 
The determination of ferrous iron was made by dissolving the 
mineral in a mixture of hydrofluoric and sulphuric acids, and 
finally titrating with KMn0 4 , the modifications of Pratt* 
being used throughout the above operation. 
The results of the analysis follow : — 

I II Average Eatios 



B 2 3 ___. 


21-21 


2136 


21-29 


•304 




•304 


Ti0 2 .... 


25-06 


24-66 


24-86 


•310 


l 


•3328 


Si0 2 .... 


1-45 


1-32 


1-39 


•0228 


MgO .... 
FeO .... 


35-41 
9-11 


36-01 
9-20 


35-71 
9-15 


•884 
•127 


1 
I 


1-011 


Fe 2 3 ... 


4-77 


4-76 


4-76 


•0297 






A1 2 3 ... 


2-95 


2-87 


2-91 


•0284 







99-96 100-18 100-07 

The amounts of sesquioxides found are comparatively small 
and the ratios obtained from them have no rational relation to 
those obtained from the percentages of the other constituents. 
Disregarding the Fe 2 3 and A1 2 3 for the present, it will be 
seen that the other constituents yield, 

B 2 3 : Ti0 2 : (Mg, Fe)0 = 1 : 1-094: 3-325. 

* This Journal (3), xlviii, 149, 1894. 



Bradley — Composition of the Mineral Warwickite. 183 

This would point to B 2 3 . Ti0 2 . 3(Mg, Fe)0 as the formula 
for the mineral. Since, however, a glassy green spinel is so 
intimately associated with the warwickite and its separation 
from it, both on account of its closely similar specific gravity, 
and because under the microscope it assumes an almost metallic 
appearance, it is thought reasonable to assign the 2*91 per cent 
of A1 2 3 found to its presence in the material analyzed. This 
assumption would necessitate the subtraction of an equivalent 
amount of MgO as required by the formula MgO . A1 2 3 . 
Qualitative and quantitative tests on this spinel have proven it 
to correspond essentially to the variety known as chlorospinel, 
in which a little of the A1 2 3 is replaced by Fe 2 3 . In this 
particular spinel there was found 8*92 per cent of Fe 2 3 , and 
corrections for this isomorphous Fe 2 3 introduced into the 
warwickite analysis by means of this spinel have been accord- 
ingly made. 

The presence of magnetite associated with the warwickite 
was proven by testing the impure material by the ordinary 
magnet. If, as is possible, it was intimately mingled with the 
warwickite it would be difficult to entirely separate it, as the 
warwickite itself is attracted easily by the electro-magnet. It 
seems reasonable therefore to assume that the greater part of 
the Fe 2 3 found was contained in magnetite, and that an 
equivalent amount of FeO to correspond to FeO . Fe 2 3 should 
be deducted from the analysis. Treating the analysis in this 
way we have the following results : 



Eatio 

1 

1-090 
3-134 













Calculated 




Average 




Spinel 


Magnetite 


III 


to 100 




BA 


21-29 






21-29 


23-87 


•341 


Ti0 2 


24-86 






24-86 


27-87 


•347 


Si0 2 


1-39 






' 1-39 


1-56 


•025 


MgO 


35'7l 


— 1-26 




34-45 


38-63 


•957 


FeO 


9-15 




— 1-95 


7-20 


8-07 


•112 


Fe 2 3 


4-76 


— '42 


— 4-34 








A1 2 3 


2-91 


— 2-91 




89-19 








100-07 


100-00 





The ratios from the corrected analysis yield B 2 3 . TiO a . 
3i Mg, Fe)0 as before, but with sharper agreement between the 
theoretical and derived numbers. The formula for warwickite 
can then be written, 

(Mg,Fe) 3 TiB 2 8 

which could be developed into a symmetrical structural for- 
mula as follows : — 



184 Bradley — Composition of the Mineral Wanoickite. 

M <o>\ 

<x o 

M < > 

The theoretical composition corresponding to this formula 
would be 

B,0 3 = 25-81 

Ti0 2 = 29-54 

3MgO = 44-65 



100-00 



In conclusion, the author here wishes to thank Professor 
W. E. Ford for his kind advice and assistance. 

Mineralogical Laboratory of the 

Sheffield Scientific School of "Yale University, 

New Haven, Conn., June, 1908. 



Chemistry and Physics. 185 



SCIENTIFIC INTELLIGENCE 



I. Chemistry and Physics. 

1. The Question of Change in Total Weight of Chemically 
Reacting Substances. — H. Landolt has devoted many years of 
the most painstaking work in investigating the weights of sub- 
stances in closed glass vessels before and after they were mixed 
to produce chemical reactions. In 1893 he published his first 
paper on the subject, in which he described work with reactions 
between silver sulphate and ferrous sulphate, iodic acid and 
hydriodic acid, iodine and sodium sulphite, and chloral hydrate 
and alkali. As a result of these researches he was unable to 
establish with certainty any change in the total weight, but it 
appeared that the separation of silver and of iodine were accom- 
panied by a slight loss in weight, so that the work was continued. 
In 1896 results were described, obtained with a balance of great 
precision, and with the use of much care, in which it was found 
that a loss was indicated, where silver or iodine were set free, in 
42 out of 54 separate experiments. The changes in weight 
resulting from the use of 60 to 120 g. of reacting mass varied 
usually between -003 and *050 mg., and were often less than the 
estimated maximum error of *03 mg. At this time it appeared 
to Landolt that these losses in weight, although small, were real, 
and he suggested the view, which was supported by the doctrine 
of the decomposition of radio-active atoms, that the violent shock 
which the atoms receive in chemical reactions might possibly 
cause the splitting off of minute particles of matter in the case of 
elements not belonging to the radio-active class, and that these 
might possess the property of penetrating the walls of glass ves- 
sels. Landolt has now published the results of additional, very 
elaborate, work on the same subject. He has explained fully 
the cause of former losses in weight in the fact that glass vessels 
which have been slightly heated by the chemical reaction within 
often do not return to their proper weight until after the lapse of 
a week or more, and he reaches the conclusion that in all of the 
15 chemical reactions studied by him, no change in the total 
weight of the reacting substances has been established. This 
conclusion is of much importance in deciding the question, 
whether or not the atomic weights are of constant magnitude, for 
there appears to be no doubt that the force of gravity acts upon 
an atom to the same extent in any state of combination. — 
Zeitschr. physikal. Chem., lxiv, 581. h. l. w. 

2. The Volume of Radium Emanation. — Professor Ruther- 
ford has recently made some measurements of the volume of the 
gaseous emanation produced by radium. For this work the 
Royal Academy of Science in Vienna put at his disposal a prepa- 
ration containing about 250 mg. of pure radium. Much diffi- 



186 Scientific Intelligence. 

culty was encountered in purifying the gas, and the results 
varied to a considerable extent. The following table gives 
examples of the amounts of emanation in equilibrium with 1 g. 
of radium : 

At the beginning At the end 

of the experiment of the experiment 

J.32C mm Q.gQCmm 

0-80 " 

0-97 " 0-66 " 

1-05 " 0-58 " 

These results, considering the difficulties encountered, agree 
fairly well with the volume, 0'b^ cmm , which Rutherford has cal- 
culated theoretically, and the lowest result is only one-ninth of 
the volume, H i'0*l cmm , found by Ramsay and Cameron some time 
ago. The gas underwent remarkable changes in volume after it 
was collected. In some instances it contracted to less than half 
its volume in the course of several hours, and then showed little 
change in the course of a week. In other cases an increase in 
volume was shown to double that of the original gas, and then a 
slow contraction followed. » These changes in volume, in many 
cases, bear no relation to the changes in volume of the emanation 
itself, for the true volume of the emanation was often only 20 
per cent of the total gas volume. The author finds no satisfac- 
tory explanation for these remarkable changes in volume. — 
Monatshefte, xxix, 995. h. l. w. 

3. A New Method for Separating Tangstic and Silicic Oxides — 
Defacqz treats the mixed acids in a boat at a red heat with a 
current of hydrogen until the tungsten is completely reduced to 
a lower oxide or to the metal. Then the boat is heated in a tube, 
so arranged as to collect the volatile products, in a current of 
perfectly dry chlorine gas. If air is absent the whole of the 
tungsten is volatilized as hexachloride and oxychloride. The 
volatile products are collected by means of ammoniacal water 
and the tungsten is determined by one of the usual methods. The 
silica in the boat is weighed after heating it in hydrogen again to 
make sure, by the absence of any blackening, that the separation 
is complete. — Bulletin, IV, iii, 892. h. l. w. 

4. A Silicide of Uranium. — Defacqz has prepared the com- 
pound Si 2 U by the aluminothermic method, using finely divided 
aluminium, flowers of sulphur, silica, and uranium oxide in proper 
proportions. The silicide forms a brilliant crystalline powder 
with metallic luster. It is interesting to notice that the author 
has prepared in the same way analogous silicides of molybdenum 
and tungsten, Si 2 Mo and Si 2 W, and that these silicides all corre- 
spond in type to the silicides of the iron group of metals having 
the general formula Si 2 M. — Comptes Hendus, cxlvii, 1050. 

h. l. w. 

5. A New Periodic Function of the Atomic Weight. — Viktor 
Poschl takes the percentage composition of the earth's crust as 



Chemistry and Physics. 187 

calculated by F. W. Clarke, and by using these percentages 
as ordinates and the corresponding atomic weights as abscissas, 
shows that there is evidence of periodicity in the abundance of 
the elements. The curve which is carried only as far as nickel 
shows four maxima at oxygen, silicon, calcium and iron, and is 
regular enough to be very suggestive and interesting. — Zeitschr. 
physikal. Chem., lxiv, 707. h. l. w. 

6. Velocity of Rontgen Ray* ; also their Influence on the 
Brush Discharge. — Erich Marx has previously measured the 
velocity of these rays (Ann. der Phys., xx, p. 677, 1906) by a 
null method, and in order to justify his use of the method he 
employed follows with another paper which is devoted to the 
theory of the method. He discusses various phenomena of the 
rays which come into prominence in his method, and naturally 
treats also of ionization. — Ann. der JPhysik, No. 1, 1909, pp. 37- 
56, 153-174. j. t. 

7. Radiation of Uranium JC. — Uranium X was obtained by 
the Moore-Schlundt and Becquerel method. Heinrich Willy 
Schmidt gives with many details the results of his investigation 
of the radiation of this substance. 

( 1 ) The hard /3-rays gave 

^ = 2-76'10 10 cm. sec -1 
€ /m=0-67'10 7 E.M.E. 

(2) The soft rays are absorbed by aluminium according to an 
exponential law, and are deviated in a magnetic field in the 
manner of negative particles. 

(3) The curves of absorption, in different substances, varies 
much with the distribution of the radiation. 

(4) An under limit is given for the absolute value of the 
reflection of the /3-rays for very thick plates. The absolute 
value of the reflected radiation is greater for the hard /3-rays 
than for the soft rays. — Physikal. Zeitschrift, Jan. 1, 1909, pp. 
6-16. j. t. 

8. Influence of Self Induction on Spark Spectra. — G. Berndt 
contends that the criticisms of Neculcea and Hemsalech on 
his investigation of the influence of self induction on spark 
spectra (Diss. Halle, 1901) do not consider the factor of time 
of exposure of the photographic plate, for the introduction 
of self induction greatly weakens the intensity of the spark. — 
Physikal. Zeitschrift, Jan. 1, 1909, pp. 28-29. j. t. 

9. Ionization of Gases, by Spark and Arc. — It is known that 
gases subjected to high temperatures in the neighborhood of 
electric sparks or the electric arc preserve their increased con- 
ductibility much longer than gases subjected to ultra-violet light, 
X-rays, or a- and /8-rays. Heinrich Rausch, in a preliminary 
paper, investigates this property with a number of gases, among 
which were ordinary lighting gas, acetylene, hydrogen, carbu- 
retted hydrogen. He found a very long persistence of conducti- 
bility in lighting gas and acetylene. — Physikal. Zeitschrift, Oct. 
25, 1908. j. t. 



188 Scientific Intelligence. 

10. Investigations in Radiation. — The recently issued number 
of the Bulletin of the Bureau of Standards (vol. v, No. 2) con- 
tains two articles on radiation to which attention should be called 
here. The first, by W. W. Coblentz, gives the result of experi- 
ments on the selective radiation from various solids. This is 
practically an examination of the emission spectra of electrical 
insulators, or transparent media, as they are called ; a line in 
which, thus far, almost no work has been done. The substances 
used were either in the form of solid rods, made in an oxy-hydro- 
gen flame, or of thick layers of the substance spread as a paste 
upon the heater of a Nernst lamp. The rods were heated by an 
electric current from the secondary of a 2000-volt 300-watt trans- 
former. The substances examined included a series of oxides, 
as those of zirconium, cerium, thorium, uranium, etc. ; also the 
minerals oligoclase, albite, orthoclase, beryl, rutile, apatite, cal- 
cite. All of these showed prominent emission bands at certain 
points; thus the oxides have a characteristic band at 2*8 to 3/a 
and a second group of bands at 4*5 to o/x, which may be due to 
the common element oxygen. The silicates have also a sharp 
emission band at 2*9/>t characteristic of Si0 2 . In the case of oligo- 
clase it is noted that the general emission, in distinction from the 
bands of selective emission, is less intense than in the other 
silicates studied. Further, the isochromatics of oligoclase are 
peculiar in that for it the emissivity is proportional to the energy 
consumed. 

A second article, by P. G. Nutting, is an important discus- 
sion of the luminous equivalent of radiation, from the standpoint 
of both objective and subjective light and with especial reference 
to the establishment of a more precise relation between light and 
its radiation, by which it can be alone measured. 



II. Geology. 

1. United States Geological Survey, Twenty-ninth Annual 
Report, 1907-1908, of the Director, George Otis Smith. Pp. 
v, 99, with two plates. — This report contains a statement of the 
work done by the various divisions of the Survey during the fis- 
cal year ending June 30, 1908. The freedom from political 
influence, the efficiency, and the high scientific esprit de corps 
which have marked the Geological Survey since its origin have 
caused it to be intrusted with various added branches of work 
which after a period of development have been organized as sepa- 
rate bureaus. This is the history of the Forest Service and the 
Reclamation Service. The technologic branch has had a more 
recent inception and the question is now under consideration by 
Congress as to the advisability of its development into a separate 
bureau of mining technology. Such a bureau would supplement, 
along purely technologic lines, the geologic work of the Survey, 
and the two bureaus could cooperate in investigations carried on 



Geology. 189 

in behalf of the mining industry. A new branch of the Survey's 
activity which has taken form during the past year is the classifi- 
cation of Government coal lands, 22,700 square miles being 
classified and valued. This work is a result of the movement for 
conservation of National resources ; a movement which in turn 
has been able to take intelligent form, as well as popularity, 
largely as a result of the more purely scientific labors of the Sur- 
vey since its organization. The work of the Survey is thus seen 
to fall into two main divisions, work of a broad scope along 
fundamental lines, whose great value and utility may only 
become widely evident after the passage of decades, and work of 
immediate utility to meet the demands of the people and of 
Congress. The Survey is to be congratulated on having per- 
sistently followed work of both divisions. 

During the year the geologic branch published 9 geologic 
folios, 1 monograph, 2 professional papers, 18 bulletins and the 
annual volume on Mineral Resources. Mention of the important 
results cannot here be made. The topographic branch mapped 
25,658 square miles, making the total area surveyed to date in 
the United States 1,051,126 square miles, or about 35 per cent. 
In Alaska 6,626 square miles were mapped, mostly on the scale 
of 1 : 250,000. Important work by the water-resources branch 
was done in the lines of stream-flow and ground-water investiga- 
tions, and investigations regarding the quality and pollutions of 
waters. 

The entire appropriation for the Survey was $1,445,020, of 
which $300,000 was expended for topographic surveys and 
$200,000 for geologic surveys. j. b. 

2. Geological /Survey of New Jersey : Henry B. Kummel, 
State Geologist, Franklin Furnace Folio. — There has recently 
been issued by the Geological Survey of New Jersey in cooperation 
with the United States Geological Survey, a geologic folio of the 
Franklin Furnace region in Sussex County. This locality is one 
of the richest mineral regions in the world, alike important 
economically for its enormous zinc deposits at Mine Hill and 
Sterling Hill, and no less scientifically for the number and variety 
of its mineral species. In addition to the zinc minerals, over 
ninety well defined species are known from this locality, and 
eleven of these have not been found elsewhere. The region 
contains also extremely valuable deposits of white crystalline 
limestone and magnetic iron-ores. 

In the descriptive text of this folio, the geography, geology and 
geologic history of this region are fully described. Complete 
information is given regarding the mineral deposits, and maps 
and cross sections show the location and shape of the valuable 
ore bodies. The folio may be obtained from the State Geologist, 
Trenton, N. J., price 25 cents, postage 15 cents additional. 

3. A Sketch of the Geography and Geology of the Himalaya 
Mountains and Tibet ; by Colonel S. G. Bueeakd, R.E., F.R.S., 
Superintendent, Trigonometrical Surveys, and H. H. Hayden, 



190 Scientific Intelligence. 

B.A., F.G-.S., Superintendent, Geological Survey of India. Pp. 
230, pis. 37. Calcutta, 1907. — The three parts of this valuable 
monograph which have thus far been issued deal comprehen- 
sively with the geography of the Himalayas and to a less extent 
with the geography of all the great mountains of central Asia. 
Part I, on "The High Peaks of Asia" (46 pp., 8 pis.), by Col. 
Burrard, gives abundant statistics as to the height and distribu- 
tion of the 75 known peaks which exceed 24,000 feet in height. 
From an interesting chapter upon errors in observations of 
altitude it appears that even in the case of the most accurately 
measured peaks the figures usually given are liable to an error of 
from 100 to 300 feet. In part II, on "The Principal Mountain 
Ranges of Asia," pp. 47-117, pis. 9-22, Col. Burrard describes in 
detail the various ranges, and shows how they originate in broad 
uplifts along axes which are generally parallel, but which often 
bifurcate or coalesce, and less frequently meet at right angles. 
A noteworthy chapter discusses observations with the plumb-line 
and pendulum which indicate that a concealed mass of excep- 
tionally heavy material lies beneath the plains of India far 
from, but parallel to, the Himalayas and their fast-growing 
subsidiary range, the Siwaliks. In part III, on "The Rivers of 
Himalaya and Tibet" (pp. 118-230, pis. 23-37), the main 
streams are classified according to both location and size. They 
are described with the same clearness and care which are given to 
the description of the peaks and ranges. Attention is frequently 
called to the marked disagreement between divides and mountain 
ranges. It is the exception for a main divide to correspond with 
a main range. The Indus river zigzags back and forth three 
times across the great Ladakh range. In the Hindu Kush region 
part of the streams flow northward across the main range ; while 
others cross it in the opposite direction flowing southward. 
Numerous other evidences indicate the young stage of the 
mountains and plateaus and the lack of adjustment of drainage to 
geologic structure. Chapters on glaciers and on recent desicca- 
tion as indicated by Tibetan lakes, complete the discussion of 
drainage. 

The monograph as a whole is not only written in a very clear and 
interesting style, but is most accurate in detail, and most care- 
fully arranged to facilitate reference. Theoretical discussions 
are not avoided, but they play a minor part and are clearly 
distinguished from accepted facts and conclusions. A valuable 
feature of the monograph is its clear statement, not only of our 
knowledge but of the limits of our knowledge of the great 
mountains of Asia. e. h. 

4. The Gases in Hocks / by R. T. Chamberlin\ Carnegie 
Institution, Washington, 1908, 8°, 80 pp. — This paper embodies 
the results arrived at by a critical study of the gases evolved by 
heating 112 specimens of rocks in a vacuum. The list includes 
all of the more important kinds of intrusive igneous rocks, lavas, 
stratified and metamorphic rocks and a few minerals. The 



Botany and Zoology. 191 

methods appear to have been well selected and carefully carried 
out. The results show that carbon dioxide and hydrogen are in 
general the gases most largely evolved, while minute amounts of 
carbon monoxide, hydrogen sulphide, methane and nitrogen are 
apt to accompany them. The author recognizes, of course, that 
a considerable, or even the larger, part of these gases were not 
contained in the rocks as such but were evolved from carbonates, 
sulphides and hydrates, and he discusses their possible origin from 
these substances and from others not known to be present in 
rocks, such as carbides and nitrides, but which might conceiv- 
ably be present in the igneous ones. Rocks, however, are such com- 
plicated bodies and the possible reactions and interactions which 
may take place at high temperatures so many and so involved 
when a large number of factors .are concerned, as brought about 
by the possible presence of sulphides, metallic oxides of a lower 
state of oxidation, carbon and even metallic particles such as 
copper and iron, that it appears possible that all of these gases 
except the nitrogen may have been produced from original solids, 
sulphides, carbonates and hydrates. While some of the gases 
such as CO Q and water vapor are undoubtedly contained in rocks 
as such, it thus becomes a matter of doubt as to how much of 
the gases evolved are to be considered original and how much 
ascribed to secondary alteration of the original minerals. In this 
connection the reviewer regrets that the work was not accom- 
panied by a microscopical examination in thin section of the 
actual specimens studied, since this would have thrown much 
light upon the presence or absence of such secondary products. 

In conclusion the author discusses the bearing of the results 
obtained upon general problems of geology and with reference 
to the early condition and origin of the earth. While no essen- 
tially new or startling facts have been brought to light by this 
undertaking, it is none the less a very useful piece of work of 
a laborious nature which has been carefully carried out and 
which will prove of service in the future in aiding to solve 
problems of chemical geology. Many investigations of just this 
character are needed before speculation upon the early history 
and character of the earth's crust can rest upon secure founda- 
tions, l. v. P. 



III. Botany and Zoology. 

1. The Forest Flora of New South Wales; by J. H. 
Maiden, Government Botanist, and Director of the Botanic 
Garden, Sydney. — This useful treatise has now begun its fourth 
volume. The training of the author for this important contribu- 
tion to science has been of a peculiar character. After having 
familiarized himself with the most approved Museum methods in 
England, he took charge of the great economic Museum in 
Sydney, where, under many discouragements, he built up a vast 



192 Scientific Intelligence. 

establishment, which has proved of immense use to the Colony 
and its sister Colonies, Early in this work of organization he 
prepared a useful treatise on the Useful Plants of Australia, 
which embodied a whole treasury of technological information. 
During this term of service he was in constant correspondence 
with all parts of Australasia, accumulating materials from all 
quarters. After the death of Baron von Mueller, Mr. Maiden 
became the Government Botanist, and he was appointed also 
Director of the Botanic Garden in Sydney, a post which the Baron 
did not occupy in his last years. 

Equipped with an unusual amount of technological information, 
Mr. Maiden has undertaken to make his Flora, as far as possible, 
practical. In this he has succeeded admirably, so that the forest 
flora is available as a hand-book even to those who are far removed 
from the southern hemisphere. The illustrations and text are of 
a high order throughout. g. l. g. 

2. Jaarboek van het Department van Langbouw in Nieder- 
landsch- Indie, 1907. — The report on the Agriculture of the Dutch 
East Indies has just come to hand. It contains a full account of 
the efficient stations in the districts, in which the more important 
technical plants and their products are studied with reference to 
improvement. The Garden at Buitenzorg and the experiment 
stations are well illustrated and described. It is no wonder that 
the Dutch have been able to maintain their place in the fierce 
competition for supremacy in the export of tropical products. 
The authorities have spared no expense or labor in applying the 
most modern methods of cultivation throughout Java and the out- 
lying islands. g. l. g. 

3. The Origin of Vertebrates ; by Walter Holbeook Gas- 
kell. 'Pp. ix + 537. London and New York, 1908 (Longmans, 
Green & Co.). — This book forms an important contribution to the 
speculation as to which particular group of existing invertebrates, 
if any, has given rise to the vertebrate animals by a process of 
evolution. For twenty years the author has held the view that 
the nervous system of the vertebrate is in part a modification of 
the alimentary canal of some invertebrate ancestor. He believes 
that the great factor in evolution has been the growth of the 
central nervous system, and that with this factor it is possible to 
trace the evolution of the mammal from the reptile, thence back to 
the amphibian and the fish ; the latter arose from the arthropod, 
and this from the annelid. The vertebrate is therefore the natural 
evolution of a primitive crustacean ancestor. This view, of 
which the author has been one of the leading exponents for many 
years, is thought to be sustained by a critical comparison of 
each organ system of tfae vertebrate with that of its supposed 
crustacean prototype. With all the possible evidence thus ably- 
presented, it will be of interest to learn whether further enthusi- 
asm will be aroused for a theory which has thus far found few 
supporters. w. k. c. 



Botany and Zoology. 193 

4. Ticks : a Monograph of the Ixodoidea ; by George H. F. 
Nuttall, Cecil Warbueton, W. F. Cooper, and L. E. Robinson. 
Part I, Argasidse. Pp. x + 104, with a bibliography of 35 addi- 
tional pages. Cambridge, 1908 (University Press). — The discov- 
ery that ticks play a most important part in the transmission of 
certain diseases of man and domestic animals, has led to a renewed 
interest in this group of parasites. The present work will contain, 
when completed, a description of all known species of the group, 
with a discussion of their structure, life history, and economic 
importance. References are made to all the important literature 
on the subject, the bibliography being printed on one side of 
thin paper, so that the titles can be cut out, if desired, and 
gummed on index cards. The work is well illustrated by half- 
tone plates and numerous text figures. w. r. c. 

5. Animal Romances ; by Graham Renshaw. Pp. 206. 
London, 1908 (Sherratt & Hughes). — A series of vivid word 
pictures of animal scenes in various portions of the world. A 
Caucasian autumn scene with its background of mountain forest, 
into which the characteristic birds, mammals, and other animals 
are projected with kaleidoscopic effect, is followed by a glimpse 
of the Malay jungle at midnight ; while the latter picture grad- 
ually dissolves into the noon-day glitter, to be in turn lost in the 
dusk of evening; a continuous procession of living creatures 
passes before the eye, each one acting its part in the full seclu- 
sion of its native haunts. Other chapters reveal the life of the 
African wilderness, the Antarctic seas, the Andean mountains, 
the Australian bush, the Pacific coral reef and other regions of 
the globe. Most of these scenes apparently have been drawn 
directly from the personal impressions of the writer, and portray 
vividly and accurately the living creature in its natural activities 
and customary environment. The illustrations are all taken from 
photographs by the author. w. r. c. 

6. Essays on Evolution, 1889-1907 ; by Edward Bagnall 
Poulton. Pp. xlviii -f- 479. Oxford, 1908 (Clarendon Press).- — 
This volume consists mainly of ten essays on the subject of 
evolution, delivered as addresses on various occasions since the 
year 1889. The text of the original essay has been altered 
whenever necessary to represent the views of the author at the 
present time, and the last and longest essay on " The Place of 
Mimicry in a Scheme of Defensive Coloration" has been entirely 
rewritten and emphasis laid on the advance in the knowledge of 
the subject in recent years. The new discoveries supporting the 
doctrines of Mendelism and of Mutation are discussed in an 
introductory chapter, and with some of the expounders of these 
doctrines the author has little patience, because of their qiiite 
unnecessary depreciation of other subjects and other workers. 
On the whole, the book forms a most interesting and important 
exposition of some of the most vital topics of Darwinian evolu- 
tion by a well known authority on the subject. w. r. c. 



191 



Scientific Intelligence. 



7. Parasitology : a supplement to the Journal of Hygiene ; 
edited by George H. F. Nutall and A. E. Shipley. — A newly 
established journal devoted to the publication of original contri- 
butions on the biology of the animal parasites of man and 
animals. 



IV. Miscellaneous Scientific Intelligence. 

1. A New Goniometer Lamp ; by Frkd. Eugene Wright. 
(Communicated from the Geophysical Laboratory.) — The gonio- 
metric measurement of minute crystal faces requires a source of 
illumination of such intensity that for the past five years the 
writer has employed an electric arc goniometer lamp (Proc. Amer. 
Philosophical Soc, xlii, 237-238, 1903) for the purpose. Certain 

Fig. 1. 



B 




f\ ' 






'HP 1 * ■ 


C 





features of the arc light, however, are not favorable for its con- 
stant use, and although, on occasion, it is the best light procura- 
blej other sources of illumination have been found better adapted 
for general work. — The Nernst light is an excellent source, and 
were it not for the present unreliability of the filaments, would 
serve every purpose. — The acetylene light, however, has been 
found by experience to be the most serviceable, and the following 
goniometer lamp has been constructed for its use. This lamp is 



Miscellaneous Intelligence. 195 

modeled in principle after the Welsbach goniometer lamp of 
Goldschmidt (Zeitschr. Kryst. xxiii, 149-151, 1894), and consists 
essentially of a tee with side outlet (tig. 1, A, l*5-in. diameter) with 
iron pipe fittings, B and C, of proper length, together with a 
base plate, D. This device fits over the acetylene burner L, and 
can be removed at any time and the aeetylene burner used for 
other purposes. The mirror, F, serves to reflect the light from 
the burner to the verniers of the goniometer, and, like the Gold- 
schmidt lamp mirror, furnishes all light requisite for goniometric 
work. By means of the brass shield plate at E, the side outlet of 
A can be opened and closed at will, and with it the light from 
the burner to the mirror. The materials of which this lamp is 
made are all on the market and can be readily procured from any 
pipe-fitting establishment and assembled at moderate cost by a 
mechanic. The acetylene burner is of the usual one-half foot 
type and the generator No. 102 of the firm of J. B. Colt, New 
York. 

2. A Containing Device for Salts Used as /Sources for Mono- 
chromatic Light ; by Fred. Eugene Wright. (Communicated 
from the Geophysical Laboratory.) — For many years sodium, 
lithium and thallium compounds have been employed to produce 
fairly monochromatic light, — yellow, red and green respectively — 





Fig. 1. 




-=^^^^^^^^^=r" — ~~— T~'~ 


r // /££= 


sE^Ei — vjp i'^__ 




-ti — - -^=gy^-^^z: ~±± — — -"- — ■ 




ij~~ == %^w s — ^ — ~ ~-— 


', 


^^s^ssX^-- -=,=.- - 



and a number of different devices for holding such salts in the 
Bunsen flame have been suggested which answer the purpose 
more or less satisfactorily. The following simple arrangement 
(fig. 1), which is apparently novel, has been found useful and 
effective by the writer in this connection and merits a brief word 
of description. The salt is placed in a small thin-walled platinum 
crucible about 1-5 to 2 cm long and 10 mm in diameter (P of fig. 1) ; 
a bundle of fine platinum wires (fig. 1, D, 4-5 cm in length serves 
as a wick and is held in proper place by pinching together one 
side of the platinum crucible, as indicated in the figure. The 
crucible is supported by a thick platinum wire L, which in turn 
is attached to the tube B of the Bunsen burner by the clamp A. 
The platinum crucible is purposely inclined at an angle as indi- 
cated in the figure, in order that its side may be reached by the 
flame and heated so hot that the salt it contains melts and is 
gradually fed into the flame by the wick of platinum wires. A 
single charge of sodium carbonate thus introduced has-been 



196 Scientific Intelligence. 

used for weeks at a time. By having on hand three such 
devices, one for sodium, the second for lithium and the third for 
thallium salt, the observer can at any instant change from the 
one to the other and proceed with his measurements for hours if 
necessary without further care for the flame. By this process of 
melting down the salts, the flow of fresh material is continuous 
and the flame is constant and remains practically unchanged for 
a long period of time. 

3. Report of the Secretary of the Smithsonian Institution for 
the year ending June SO, 1908. Pp. 84. Washington, 1908. —The 
annual report of Dr. Charles D. Walcott, Secretary of the Smith- 
sonian Institution, has recently been issued. It gives the usual 
interesting account of the workings of the Institution in its varied 
functions, prominent among which are the National Museum, the 
Bureau of American Ethnology, the International Exchanges^ the 
National Zoological Park, and the Astrophysical Observatory. 
The work of the Institution is now so well organized that it goes 
forward in a manner most satisfactory to all the interests involved. 
In regard to the new building for the National Museum, it is 
stated that the walls are completed and the construction of the 
roof well under way. .There remains, however, the fitting up to 
the interior, including some ten acres of floor space. The most 
interesting part of the work of the Bureau of Ethnology has been 
the excavation and repair of the Casa Grande ruins in Arizona, 
under the charge of Dr. Fewkes. Although the work has not 
been completed it has progressed far enough to present a typical 
ruin, given the general character of the ancient Pueblo remains 
of that region. This most interesting subject, and others related, 
are described in detail in an Appendix to the present Report, 
prepared by the Chief of the Bureau, W. H. Holmes. Other 
appendixes are given by the gentlemen in charge of the different 
departments, among which must be mentioned that by C. G. 
Abbot on the work of the Astrophysical Observatory. A number 
of special investigations are enumerated which are now being 
carried on by grants from the funds of the Institution. 

The Annual Report of the Board of Regents for the year end- 
ing June SO, 1907, has also been issued. This contains the Report 
of the Secretary, issued in advance about a year since (see vol. 
xxiv, p. 160). There is also the usual Appendix, pp. 95-709, con- 
taining selected articles of general scientific interest on a wide 
range of topics. As most of these are not easily accessible in the 
original, their republication here should be of great value to the 
intelligent reading public. 

Obituary. 

George W. Hough, Professor of Astronomy at Northwestern 
University and Director of the Dearborn Observatory, died at 
his home in Evanston on January 1 in his seventy-third year. 



New Circulars. 



84 : Eighth Mineral List : A descriptive list of new arrivals, 

rare and showy minerals. 

85 : Minerals for Sale by Weight : Price list of minerals for 

blowpipe and laboratory work. 

86 : Minerals and Rocks for Working Collections : List of 
common minerals and rocks for study specimens ; prices 
from ij4> cents up. 

Catalogue 26: Biological Supplies: New illustrated price list 
of material for dissection ; study and display specimens ; 
special dissections; models, etc. Sixth edition. 

Any or all of the above lists will be sent free on request. We are' 
constantly acquiring new material and publishing new lists. It pays to 
be on our mailing list. 



Ward's Natural Science Establishment 

76-104 College Aye., Rochester, N. Y. 

Ward's Natural Science Establishment 

A Supply-House for Scientific Material. 

Founded 1862. Incorporated 1890. 

DEPARTMENTS : 

Geology, including Phenomenal and Physiographic. 
Mineralogy, including also Eocks, Meteorites, etc. 
Palaeontology. Archaeology and Ethnology. 

Invertebrates, including Biology, Conchology, etc. 
Zoology, including Osteology and Taxidermy. 
Human Anatomy, including Craniology, Odontology, etc. 
Models, Plaster Casts and Wall-Charts in all departments. 



Circulars in any department free on request ; address 

Ward's Natural Science Establishment, 

76-104 College Ave., Rochester, New York, U. S. A. 



CONTENTS. 

Page 
Art. VII — Revision of the Protostegidse ; by G. R. Wieland. 

(With Plates II-1Y) 101 

VIII. — Submarine Eruptions of, 1831 and 1891 near Pantel- 

leria ; b}^ H. S. Washington . _ 131 

IX. — Types of Permian Insects ; by E. H. Sellards ._ 151 

X. — lodometric Estimation of Vanadic Acid, Chromic Acid 

and Iron in the Presence of One Another ; by G. Edgar 174 

XI. — Analysis and Chemical Composition of the Mineral 

Warwickite ; by W. M. Bradley _ . . 179 

SCIENTIFIC INTELLIGENCE. 

Chemistry and Physics — Question of Change in Total Weight of Chemically 
Eeacting Substances, H. Landolt : Volume of Eadium Emanation, 
Rutherford, 185. — New Method for Separating Tungstic and Silicic 
Oxides, Defacqz: Silicide of Uranium, Defacqz: New Periodic Func- 
tion of the Atomic Weight, V. Poschl, 186. — Velocity of Eontgen Rays ; 
also their Influence on the Brush Discharge, E. Marx : Eadiation of 
Uranium X, H. W. Schmidt : Influence of Self Induction on Spark 
Spectra, G. Berndt : Ionization of Gases by Spark and Arc, H. Eausch, 
187. — Investigations in Eadiation, W. W. Coblentz and P. G. Nutting, 
188. 

Geology — Twenty-ninth Annual Eeport United States Geological Survey, 
G. 0. Smith, 188. — Geological Survey of New Jersey, H. B. Kummel : 
Sketch of the Geography and Geologj*- of the Himalaya Mountains and 
Tibet, S. G. Burrard and H. H. Hayden, 189. — Gases in Eocks, E. T. 
Ckamberlin, 190. 

Botany and Zoology — Forest Flora of New South Wales, J. H. Maiden, 
191 — Jaarboek van het Department van Langbouw in Niederlandsch- 
Indie, 1907: Origin of Vertebrates, W. H. Gaskell, 192.— Ticks: a 
Monograph of the Ixodoidea, G. H. F. Nuttall, C. Warburton, W. 
F. Cooper, and L. E. Eobinson : Animal Eomances, G. Eenshaw : Essays 
on Evolution, E. B. Poulton, 193. — Parasitology, G. H. F. Nutall and 
A. E. Shipley, 194. 

Miscellaneous Scientific Intelligence— New Goniometer Lamp, F. E. Wright, 
194.— Containing Device for Salts Used as Sources for Monochromatic 
Light, F. E. Wright, 195. — Eeport of the Secretary of the Smithsonian 
Institution for the year ending June 30, 1908, 196. 

Obituary — G. W. Hough, 196. 



Dr. Cyrus Adler, 

Librarian U. S. Nat. Museum. 






VOL. XXVII. 



MARCH, 1909. 



Established by BENJAMIN SILLIMAN in 1818. 
THE 

AMERICAN 

JOURNAL OF SCIENCE. 

Editor: EDWAKD S. DANA. 

ASSOCIATE EDITORS 

Professors GEOKGE L. GOODALE, JOHN TROWBRIDGE, 
W. G. FARLOW and WM. M. DAVIS, of Cambridge, 

Professors ADDISON E. VERRILL, HORACE L. WELLS, 
L. V. PIRSSON and H. E. GREGORY, of New Haven, 

Professor GEORGE F. BARKER, of Philadelphia, 
Professor HENRY S. WILLIAMS, of Ithaca, 
Professor JOSEPH S. AMES, of Baltimore, 
Mr. J. S. DILLER, of Washington. 



FOURTH SERIES 

No. 159— MARCH, 1909. 

VOL. XXVII-[WHOLE NUMBER, CLXXVIL] 






NEW HAVEN, CONNECTICUT. 
1909. 



THE TUTTLE, MOREHOUSE & TAYLOR CO., PRINTERS, 123 TEMPLE STREET. 




Published monthly. Six dollars per year, in advance. $6.40 to countries in the 
Postal Union ; $6.25 to Canada. Remittances should be made either by money orders, 
registered letters, or bank checks (preferably on New York banks). 



EAEE CINNABARS FROM CHINA. 

We would call attention to these remarkable Cinnabars, some of which 
still remain. They are beyond doubt the most beautiful and interesting 
crystals of Cinnabar ever discovered. They were described and illustrated 
in this Journal, Nov. '08. The prices range from $5, $7.50, $10, $15, $18, 
$25, $35, and $50. 

NEW ARRIVALS. 

Euclase, Capo do Lane, Brazil; Chalcocite, Conn.; Columbite, Conn.; 
Monazite, large loose xls. and in matrix, Conn.; Uraninite, crystal in 
matrix, Conn.; Benitoite, San Benito Co., Cal.; Neptunite, Cal.; Lievrite, 
Elba ; Polybasite, Hungary and Mexico ; Herderite, Poland, Maine ; Cali- 
f ornite, Tulare Co. , Cal. ; Cobaltite, loose crystals and in matrix, Cobalt, 
Ont., and Sweden; Vivianite, large crystals, Colo.; Olivinite, Utah ; Sarto- 
rite, Canton Wallis ; Jordanite, Binnenthal ; Crocoite, Siberia and Tasmania ; 
Cinnabar, Cal., Hung, and China; Gypsum, twin crystals, Thuringia ; 
Diamond in matrix, New Vaal River Mine, South Africa ; Argentite, Mexico ; 
Freiberg, Saxony ; Pyrargyrite, Saxony and Mexico ; Pyromorphite, Ems, 
Germany, Phoenixville, Pa. ; Tourmalines, beautiful sections from Brazil ; 
Brochantite, on Chrysocolla, Utah ; Pink Beryl, small and large, Mesa 
Grande, Cal.; Kunzite, small and large, Pala, Cal.; Sphene, Binnenthal; 
Titantite, Tilly Foster, N. Y. ; Tetrahedrite, Utah and Hungary; Realgar, 
Hungary ; Opal, Caribou River, Queensland ; Torbenite, Eng. ; Bismuth, 
native, Cobalt, Ont.; Emerald, loose and in matrix, Ural ; Zircon crystals, 
loose, Ural ; Green and Cinnamon Garnets, Minot, Maine ; Vesuvianite, 
Poland, Maine, Italy and Tyrol ; Zeolites, beautiful specimens from Erie 
Tunnel, Patterson, and Great Notch. 



NORWAY and SWEDISH MINERALS. 

Just secured a fine lot of exceedingly rare minerals ; as they are now in 
the Custom House, and cannot be listed in time for this issue, kindly write 
for list of same. 



CUT GEMS. 



Garnets, green and red ; Aquamarines ; Zircons, all shades ; Sapphires, 
all shades ; Star Sapphires and Star Rubies ; Chrysoberyl, Cats-eye ; Spinels, 
all shades ; Topaz, pink, blue, brownish and golden color ; Pink Beryl ; 
Sphene ; Tourmaline, all shades ; Amethyst, Siberia, royal purple Star 
Quartz ; Peridote ; Opal matrix, Mexico and Australia ; Hyacinth ; Tur- 
quoise, Mexico and Persia : Kunzite ; Reconstructed Rubies and Sapphires ; 
Opal carvings ; Antique Cameos and Mosaic, and other semi-precious stones. 

Let us know your wants, and we will send the specimens on approval to 
you. Write for our new circular to-day. 



A. H. PETEREIT, 

8i — 83 Fulton Street, New York City. 



THE 



AMERICAN JOURNALOFSCIENCE 

[FOURTH SERIES.] 



Art. XII. — Recent Observations in Atmospheric Electric- 
ity ;*by P. H. Dike. 

It is somewhat difficult to give a satisfactory resume of recent 
work in Atmospheric Electricity which can claim to be complete 
and at the same time strictly up to date, since aside from the 
four or five centers from which emanate most of the contribu- 
tions to the advance of the subject, work is frequently coming 
to light from unexpected sources. 

We have been in the habit of looking to Elster and Geitel, 
Gerdien, Ebert, and a few others in Germany, to the Caven- 
dish Laboratory in England, and to Rutherford and his fol- 
lowers at McGill University, Canada, for most of the work in 
Atmospheric Electricity, but of late other experimenters have 
begun to enter the field. 

Sot very much has been done as yet in this country along 
these lines, and what has been done has been scattering and 
desultory for the most part, consisting of the work of an indi- 
vidual here and there. There has been no such extended and 
persistent study of the subject, and such careful experimenta- 
tion as has been going on in England and Germany for some 
years, and which has yielded so many important and interest- 
ing results. For this reason most of the work referred to will 
necessarily have been done in foreign laboratories. The field 
of research in Atmospheric Electricity is a broad one and pre- 
sents numerous complications of a most troublesome sort, 
which offer to the observer abundant opportunity for the 
exercise of his ingenuity. 

The earliest known and most thoroughly studied of all the 
phenomena included under the head of Atmospheric Electricity 

* Presented before the Philosophical Society of Washington, October 24, 
1908. 

Am. Jour. Sci.— Fourth Series, Vol. XXVII, No. 159.— March, 1909. 
14 



19S Dike — Recent Observations in Atmospheric Electricity. 

is the potential gradient — the existence of an electric field about 
the earth, resulting in a difference of potential between the 
earth and an} T point in the air above it. The existence of such 
a potential is readily shown by the use of a suitable electro- 
scope or electrometer connected to a collector, such as a flame, 
a water dropper or a radio-active body. Potential gradients 
exceeding 100 volts/meter are ordinary, and during times of 
electrical disturbance, even with a clear sky, this may increase 
to 1000 volts or more. The mean value at Kew, England, has 
for eight years exceeded 159 volts per meter, while on January 
1, 1908, during a fog it exceeded 730 continuously for 8-J hours. 
Its fluctuations are abrupt and of great range, sometimes passing 
suddenly from large positive to negative values. At Kew, at 
Potsdam, and at various other observatories, continuous records 
have been kept of this most erratic phenomenon by means of 
self-recording instruments for years, in an attempt to establish 
some relationship between it and other meteorological elements, 
but so far without very definite results, except for a possible 
connection with the barometric pressure ; and this relationship, 
according to Chree, seems to point to change of potential 
gradient as a cause, and change of pressure as a result. It 
would seem to be almost impossible to correlate such a phenome- 
non with local weather conditions except in so far as electri- 
cal storms are concerned, when it is considered that as a charged 
body we have the earth, whose electrical condition is depend- 
ent on conditions prevailing over its whole surface rather than 
on local phenomena. 

By recent observers the study of this element of the problem 
has been made subsidiary to other investigations in which it is 
involved as a factor, principally in the computation of the 
earth-air currents. 

An interesting series of investigations with a practical end 
in view has been made by Creighton of the General Electric 
Company bearing upon the subject of electrical storms with 
especial reference to lightning discharges and the operation of 
lightning arresters on transmission lines. Laboratory experi- 
ments were supplemented by observations at two power plants 
in Colorado where electric storms are of daily occurrence in 
summer. By means of a rotating film or a moving sensitive 
tape the duration of a discharge on a transmission line was 
measured by means of projecting the light from a spark gap 
connected with the line upon the film. A range of duration 
of from 0*04 sec. to 0*0001 sec. was found in 36 discharges, the 
greater number being of very short duration and only two 
exceeding 0*01 sec. As many as seven discharges in a second 
were observed. The discharge is oscillatory, showing 3000 
cycles per second, corresponding to the line frequency, and 



Dike — Recent Observations in Atmospheric Electricity. 199 

may be further analyzed into higher frequencies of the order 
of 1,000,000 cycles per second by coil resonators arranged to 
affect strips of sensitive paper by their brush discharge when 
excited by frequencies of the proper pitch. The potentials were 
approximately measured by graduated needle gaps, a number of 
which were arranged in parallel rows, with fuses and high 
resistance in series, the potential falling between the minimum 
sparking pressure of the longest gap crossed and that of the 
next higher. The measurement of the lightning current was 
difficult, as it depends on the capacity, inductance and resist- 
ance of the lines and the frequency of discharge. The 
quantity of the discharge over the line was estimated by not- 
ing the size of fuse that it would just burn out, several fuse 
wires of different sizes being placed in series, the arc holding 
in the vapor of the first fuses blown till the larger ones were 
melted. Using the data from one of these tests, the quantity 
of discharge was estimated at 0*003 of a coulomb, and if this 
quantity of electricity was spread out over a mile of wire 
having a capacity between line and ground of 0*01 micro-farad 
the initially impressed voltage would be about 600,000 volts. 

The development of the theory of the ionization of gases 
opened up a new field of investigation in Atmospheric Elec- 
tricity, since it afforded an explanation of the previously 
observed fact, that a charged body exposed to the air lost its 
charge at a more rapid rate than could be explained by leak- 
age over the insulating supports. Elster and Geitel were pion- 
eers in this field, and though the instruments devised by them 
for observing the "dispersion" gave erroneous results through 
the establishment of saturation currents, that is, currents inde- 
pendent of the potential difference, once a certain minimum 
potential is passed, they accumulated much information of a 
relative nature. It is to J. J. Thomson and his school that 
we are indebted for a better understanding of the conductivity 
of gases, leading to the construction of apparatus on more 
correct principles ; results are now being obtained which may 
probably pretend to some degree of accuracy. The experimen- 
ter in these lines must be satisfied with somewhat approximate 
results, since it is obviously impossible to measure a con- 
ductivity to one or two per cent when its value is varying dur- 
ing the time of measurement by perhaps fifty per cent. The 
order of magnitude can be arrived at, but the conductivity of 
a gas is quite a different matter to that of a metal where the 
number of ions seems to be infinite, and atmospheric air is even 
more unstable in its properties than a confined sample of gas. 

The establishment of saturation currents in the apparatus for 
determining the conductivity of the air may be prevented in 
two wavs, either by the removal of all other conductors to such 



200 Dike — Recent Observations in Atmospheric Electricity, 

a distance from the charged body, whose loss of charge is to be 
measured, that at the potential to be used the currents shall be 
unsaturated, or by setting up a current of air about the charged 
body of such velocity that the same result is reached. 

The latter, first applied in its fullest extent by Gerdien, has 
proved to be of great use for field work, while the former 
seems more suitable for observatory use, where continuous 
records are desired. The Gerdien conductivity apparatus, 
briefly described, consists of a cylindrical condenser, the outer 
cylinder of which is 16 cm in diameter and 35 cm long, while the 
inner cylinder is l*4 cm in diameter and 24 cra long, and is con- 
nected with an electroscope by means of which the potential 
of a charge put upon the cylinder can be observed. A current 
of air, of sufficient velocity to prevent the establishment of 
saturation currents, is drawn through the condenser with the 
aid of a fan, thus using the second of the two methods of avoid- 
ing saturation. This instrument has been used by various 
observers on land, in balloons and at sea, with very consistent 
results, and it has proved very satisfactory for field work. 
Gerdien has made use of it at Gottingen in conjunction with 
a self-recording instrument giving the course of the potential 
gradient for determining the value of the earth-air current. 

The potential gradient and the conductivity being known, 
the current per square centimeter of the earth's surface i& 
readily computed. 

bv /x x . 

A series of observations carried out during about two and 
a half weeks in April, 1906, gave as a mean from 49 measure- 
ments 8 X 10~ 7 electrostatic units per square cm. for the 
value of the vertical current, or 2'7xlO~ 1G amperes. The con- 
ductivity was found to be X / =1'155 X 10~ 4 and X w =l*120 X 
10" 4 electrostatic units. 

I have made use of the same type of instrument at sea, on 
board the Galilee during the cruise recently completed, begin- 
ning at Sitka and ending at San Francisco, obtaining observa- 
tions, mostly in fair weather conditions, in latitudes ranging 
from 55° 41/ North to 45° 07' South, crossing the tropics twice, 
thus securing a wide range of climatic conditions. Unfortu- 
nately on board ship I was unable to measure the normal 
potential gradient in an undisturbed field, and the disturbances 
were so great and so variable that it seemed impossible to avoid 
or to correct for them. Hence my observations give only the 
conductivity and the potential gradient can only be surmised. 
A few observations made in a skiff at sea in very calm weather 



Dike — Recent Observations in Atmospheric Electricity. 201 

gave values about the same as on iand, and I see no reason for 
supposing that the potential gradient should be less over the 
sea than over the land. The conductivities obtained varied 
considerably but not more so than at a hxed land station, 
apparently exhibiting some tendency to decrease as the baro- 
metric pressure increases. 

The mean conductivities for the voyage were : 

Ap=l'603 X 10 -4 electrostatic units 
\ n — 1-433 X10- 4 " " 

This would make the vertical current somewhat greater 
than what Gerdien found, assuming the same potential gradient. 
The observations for a day usually extended over an hour or 
more and were made up of alternate readings of the conduc- 
tivity for positive and for negative electricity, A charge of 
about 150 to 200 volts was put on the inner cylinder by means 
of a dry pile and the divergence of the electroscope leaves 
read. This reading was the most difficult part of the observa- 
tion at sea since the rolling of the ship gave the leaves a pen- 
clular motion, while the rise and fall of the ship as it rode over 
the waves gave them a flapping motion, so that they alter- 
nately spread farther apart as the ship sank into the trough of 
the waves and fell together as it rose to the crest. The pen- 
clular motion was avoided to some extent by mounting the 
instrument on a gimbal stand and steadying it with the hands 
to keep it horizontal, but the flapping was unavoidable and 
limited the observations to moderately smooth seas. Only one 
leaf at a time could be watched through the eye- piece, so that 
the estimation of the mean position of the leaves was difficult. 
However, the consistency of the results shows that some degree 
of accuracy was obtained. 

Immediately after reading the divergence of the leaves, the 
fan was rotated at a uniform rate for five minutes. At the 
end of that time a second reading of the divergence was made 
and the charge was removed by earthing the inner cylinder, 
which was then charged negatively. After a wait of about one 
minute to allow any of the former charge which might have 
been absorbed by the insulation to be neutralized, a reading of 
the divergence of the leaves for the negative charge was made, 
in time to begin the rotation of the fan two minutes after the 
end of the rotation for the positive charge. Closing both ends 
of the cylinder to exclude air currents, a leakage test to deter- 
mine the normal rate of leak over the insulation was made, 
usually extending over fifteen to twenty minutes. 

The electroscope had to be calibrated of course, and this 
was done at long intervals, as laboratories for testing electrical 
instruments are not numerous in the Pacific. An apparatus 



202 Dike — Recent Observations in Atmospheric Electricity. 

was carried designed to give a rough calibration on board ship, 
but its results, after a test at Sitka, and another, an unsuccessful 
one, at Honolulu, were judged too rough to be relied upon. 
It was possible to obtain a good calibration at Christchurch, 
New Zealand, in the electrical laboratory of Canterbury College 
in January of 1908, but there was no further opportunity 
until the return of the instruments to this city in July, when 
a calibration was carried out at the Bureau of Standards, agree- 
ing remarkably closely with the curve obtained at Christchurch, 
showing that at least during the last half of the voyage the 
values had remained practically constant, that is within three 
per cent, which is well within the accuracy of reading at sea. 
A comparison of these values with those obtained at Sitka 
indicates a gradual increase in sensitiveness, which goes on 
at a diminished rate between Christchurch and Washington. 
Through unexpected good fortune I was able to use the same 
pair of leaves throughout the voyage uninjured, though the 
electroscope had to be opened to clean the insulation, thus 
exposing the leaves to air currents and other dangers of damage, 
nearly every time the instrument was used. 

An electroscope is far from being a satisfactory instrument 
for use at sea, but it is the only form of electrometer now avail- 
able that would work at all, unless some form of the "string 
electrometer" may be found satisfactory. 

C. T. R. Wilson of the Cavendish Laboratory has made some 
interesting observations of the earth-air current with a very 
different instrument, designed by himself, by means of which 
he is able to measure the current from a test plate which is 
kept at zero potential, that is, at the potential of the surface. of 
the earth, thus approximating closely to the actual conditions. 

The case of a gold-leaf electrometer is kept at a constant 
potential by means of a quartz Leyden jar, while the gold leaf 
is attached to the rod bearing the test plate, a blackened brass 
disk 7 cm in diameter surrounded by an earth-connected 
guard ring. A cylindrical cover rests upon the guard ring 
and shields the test plate except when it is to be exposed 
to the influence of the earth's electrical held. During exposure 
the test plate and gold leaf are maintained at zero potential by 
means of a compensator consisting of a cylindrical condenser 
of which the inner conductor is a metal rod connected to the 
gold-leaf system, the outer condenser being a brass tube main- 
tained during any series of observations at a constant negative 
potential by means of a quartz Leyden jar and capable of sliding 
parallel to its length to give a variable capacity. A scale is 
attached to give the position of the outer cylinder relative to 
the inner and this scale is calibrated to determine the change 
of capacity corresponding to one scale division. Thus the 






Dike — Recent Observations in Atmospheric Electricity. 203 

change of charge, that is the loss or dispersion of electricity 
from the exposed test plate, can be measured by reading the 
scale before and after exposure, the gold leaf being kept con- 
stantly at its zero position by means of a gradual motion of 
the compensator. 

The test plate being covered, the gold-leaf system is earthed 
and the position of the compensator is read. The earth con- 
nection is then broken and the cover removed. The observer 
then keeps the gold leaf at its zero position, that is, keeps the 
test plate at zero potential by means of the compensator till a 
measured interval of time has elapsed, when the position of 
the compensator is again read. From the difference of the 
two readings the loss of charge can be computed and thus the 
current per square cm. of the test plate determined. Alternate 
measurements were made with the bare test plate and with a 
piece of turf covering the plate to represent a section taken 
from the surface of the earth. The dissipation factor is 
expressed as the percentage of the charge on an earth-connected, 
exposed body which is neutralized per minute. The dissi- 
pation factor for the plate alone was found to be practically 
the same as for the turf, though the charge on the turf was 
about three times as great and the distribution of charge very 
different. The agreement is sufficiently good to warrant assum- 
ing a definite dissipation factor depending on the condition of 
the atmosphere. In order to get a closer approximation to 
actual conditions at the surface of the ground, instead of at the 
elevated test plate, a large wooden test plate and guard ring 
with their surface very little above the surface of the ground 
were constructed and connected with the test plate of the appar- 
atus. Alternate determinations were made with the small and 
the large test plates, precautions being taken that the field 
should not be disturbed by the observer. From those observa- 
tions it was found that the mean density of electrification upon 
the exposed test plate when at zero potential was nearly 4*2 
times that upon the surface of the ground. Using this reduc- 
tion factor, observations extending over a year give a mean 
value for the current per square cm. of ground 2'2XlO~ 16 am- 
peres, agreeing with Gerdien's value 2'7X 10" 16 amperes deduced 
from measurements of conductivity and potential gradient. 
The dissipation factor was least for cloudless, calm days, great- 
est for days with cumulus and clear atmosphere, and inter- 
mediate for overcast days. The agreement between the two 
widely different methods is doubtless partly accidental, but is 
none the less gratifying as indicating that definite results are 
being obtained which can be duplicated by different observers 
using different methods. 



204 Dike — Recent Observations in Atmospheric Electricity. 

Sobering at Gottingen lias worked on the problem of a self- 
recording apparatus for the measurement of specific conduc- 
tivity, avoiding saturation currents by removing the charged 
body from the neighborhood of any earthed conductor. A 
suspended sphere of 5 cm radius connected by a fine wire 50 cm 
long to an electroscope was found to answer this requirement 
and to give results following Ohm's law and independent of 
air currents. This form of apparatus is not convenient for 
field work but may be made useful for observatory work by 
using a long tbin wire in place of the sphere. This must be 
carefully screened from the earth's field by surrounding it with 
a cylinder of wire netting. A sort of arbor of wire netting 
20 cm long was constructed at the Gottingen Geophysical Insti- 
tute in the open and a wire 0'14 mm in diameter was stretched 
inside this so as to be at least # 5 m from the netting, and sup- 
ported by specially constructed insulators to make possible its 
continuous use out of doors. One end was connected with an 
electrometer indoors. Comparisons between the sphere disper- 
sion apparatus inside the screen and Gerdien's conducting 
apparatus outside gave the same difference as with both instru- 
ments outside. The capacity of the wire was found to be 104 cm 
and with the normal conductivity of the air with a potential 
of 100 volts on the wire the current should be 1 X 10 " amperes. 
This being too small to measure with a galvanometer, a 
uranium cell was used as a high resistance to adapt the arrange- 
ment to continuous registration by electrometer methods. The 
uranium cell gave a saturation current, hence a constant cur- 
rent, from 50 to 300 volts and was arranged so as to be adjust- 
able to give different currents by screening some of the uranium 
deposit. If the inner electrode of the uranium cell is connected 
with the dispersion wire and the outer electrode with a poten- 
tial of some hundred volts, the inner electrode and wire will be 
charged to the potential at which the current from the wire into 
the air is equal to the current in the uranium cell. If V is 
this potential and Z the capacity of the effective portion of 
the wire and the current in the uranium cell is i, the specific 

i 
unipolar conductivity of the air is \= . 

The value of i being constant, it is only necessary to record 
"Fto follow the course of the conductivity. This was done 
photographically by means of the electrometer, the wire being 
connected with the needle and the quadrants charged by means 
of a storage battery of 20 elements. In the registration- 
various difficulties were encountered, mainly in failure of insula- 
tion of tbe wire through moisture or spider webs. 

The apparatus charges slowly through the uranium cell after 
being discharged, so that it is only suitable for slow registra- 



Dike — Recent Observations in Atmospheric Electricity. 205 

tion and not for recording rapid changes in the conductivity. 
For the latter purpose the same wire was used later with an 
electrometer with a sensitiveness of 0*82 cm per volt and the 
zero so chosen that with a charge of 88 volts the spot of light 
fell on the paper. A clock was arranged so that at minute 
intervals the wire was charged to 88 volts and then insulated. 
It then discharged itself for a minute to a certain point, the 
electrometer tracing an oblique line on the paper, the lower 
end of which gave the fall in potential in one minute. The 
trace of these points gives the conductivity curve. The curves 
given show large and rapid variations of the conductivity, 
apparently indicating that the air is not uniformly conducting 
but varies rapidly from place to place so that the wind brings 
air of different conductivities in contact with the wire. The 
second method of registration is recommended for eclipse work 
since it gives the mean values for shorter intervals than the 
usual apparatus. 

Another subject of great interest to students of atmospheric 
electricity is that of the radio-active emanation in the air, first 
observed there by Elster and Geitel. Closely related to this is 
the radio-activity of soils and rocks, sea water, etc., as the prob- 
able source of the emanation. It cannot be said that it has 
been proved conclusively what the source of this emanation is, 
though it seems to be associated with land and to be almost 
entirely absent over the sea, making it seem probable that its 
source is in the soil and rocks of the land areas since sea water 
is only j-^q-q as radio-active as the average sedimentary rock. 
Several observers have been at work on the problem of the 
determination of the quantity of the emanation per cubic meter 
of the air, or what amounts to the same thing, the determina- 
tion of the radium equivalent per cubic meter, that is the 
amount of radium required to maintain in equilibrium the 
quantity of emanation found in the air. The electrical method 
by which the emanation was first collected is not adapted to 
these measurements, since some of the carriers of the emanation 
are very slow moving and are almost sure to get past the nega- 
tively charged collector. Also some of the emanation appears 
to be attached to negatively charged particles and would thus be 
repelled from the collector, while there are also neutral carriers 
resulting from the union of a positive and negative particle. 

However, there are at least two other methods of collecting 
the emanation which seem to be quite efficient. One is by 
means of cooling a current of air in a condenser surrounded 
with liquid air, the emanation being condensed and remaining 
behind. The other is by the use of pulverized cocoanut char- 
coal, which, on allowing a current of air to pass through it, 
absorbs the emanation contained in it and gives it up on being 



206 Dike — Recent Observations in Atmospheric Electricity. 

heated to redness. Both methods were tested by Satterly at 
the Cavendish Laboratory and careful series of observations 
were made. Using the absorption method, the pulverized char- 
coal was packed in porcelain tubes between asbestos plugs and 
the air drawn through at a constant, measured velocity by 
means of a filter pump, beiug freed from dust and dried by 
means of calcium chloride before reaching the charcoal. The 
air current was measured by a pressure gauge. After the air 
had passed through for a known interval of time (the volume 
per minute being known) the tube was placed in a furnace and 
heated to redness. Air freed from emanation was passed 
through the tube while hot, sweeping out the released emana- 
tion, and was collected in aspirators over water, where it was 
stored till ready for the test. The testing vessel was in the 
meantime tested for air and insulation leaks, the air being 
pumped out of the vessel and fresh air allowed to enter several 
times. The air was finally exhausted from the testing vessel 
and communication made w T ith the aspirator containing the air 
loaded with emanation, which then passed into the testing 
vessel. The leakage current was measured for 20 minutes, 
not waiting for the 3-hour maximum. 

In order to refer these results to radium as a standard, air 
was bubbled through a solution containing a known amount 
of radium, thus removing the emanation as it was given off. 
This went on simultaneously with the collection of emanation 
from the air, a similar charcoal tube being used for absorbing 
the radium emanation. This was treated in the same way as 
the other and a similar leakage test made. This second test 
gave of course the radium emanation plus the emanation from 
the air, but since the observations were simultaneous the air 
effect could be allowed for. The comparison gives the ratio 
oi; the emanation in a known volume of air to the emanation 
generated in the solution in a known time. The results must 
be corrected for decay of the emanation if not measured for 
some time after it was absorbed, as was sometimes the case, and 
for the gradual accumulation of emanation in the charcoal. 
The amount of emanation generated in a given time by the 
solution must be calculated. The resulting mean from eight 
measurements gives as the amount of radium required per 
cubic meter of air to keep up the quantity of emanation to the 
observed amount 88X10 -12 grams with a range from 50 XlO -12 
to 160 XlO -12 . The comparison of day with night values seem 
to indicate- more emanation during the day than at night. 

For the liquid air condenser method, a special condenser was 
devised with a large cooling surface and small resistance to the 
air current, consisting of a brass tube packed as full as possible 
with thin straight brass wires. This was immersed in liquid 



Dike — Recent Observations in Atmospheric Electricity. 207 

air and required 3/4 liter of liquid air for a three-hour run at 
the rate of 1/2 liter of air per minute. The tests were very 
similar to those in the absorption method, except that on 
account of the large amount of liquid air required simultaneous 
tests with the radium solution could not be made, so they had 
to be alternated. A first series of seven measurements gave 
140 XlO -12 grams of radium as the equivalent, and a second 
series of three gave 90xl0~ 12 grams. A third series, using in 
one case sulphuric acid and in another a freezing mixture as 
dehydrators, gave 130 XlO" 1 ' 2 grams. 

Parallel tests made with the two methods — condenser and 
charcoal — show the condenser to be rather more efficient than 
the charcoal. It does not seem likely that the radium emana- 
tion and its products in the air are responsible for anything 
like all the natural ionization of the air. Eve, in Canada, has 
used the charcoal absorption method of measuring the radium 
equivalent per cubic meter and found as the mean of obser- 
vations extending over a year 60 XlO -12 grams. The ratio of 
the maximum to the minimum values found is seven to one. 
Summer and winter values are about the same. Deep cyclones 
with heavy rain cause increase while anti-cyclones cause a 
decrease in the value. Ashman, working at the University of 
Chicago, used the liquid air method, comparing the rate of dis- 
charge produced by the emanation from a known volume of air 
with that produced by the radiation from a portion of a mineral 
containing a known amount of uranium, assuming the amount 
of radium associated with one gram of uranium to be 5*4 xlO" 7 
grams. 

The result of four tests gave as the radium equivalent 
97 X 10 ~ 12 grams. Experiment showed that the coil condensed 
all the emanation and that two similar coils in parallel gave 
duplicate results. The highest value, 200 XlO" 12 grams, was 
obtained after a heavy rain and general thaw following several 
weeks of freezing weather with the ground covered with snow. 
The activity curve of the emanation showed it to be identical 
with that of radium. The values found by all three observers 
are thus seen to be of the same order, and in as good agree- 
ment as could be expected considering the variability -of the 
quantity to be measured. 

G-erdien has made an elaborate study of the velocities of 
the carriers of the emanation in the atmosphere, using the 
trajectories of the particles carried by a uniform current 
of air and acted upon by an electrical field about a cylindrical 
conductor whose axis is parallel to the direction of the air cur- 
rent, as a basis of computation. The deposit on the cylinder 
was collected, and according to its distribution over the electrode 
and according to its rate of decay was differentiated into car- 



208 Dike — Recent Observations in Atmospheric Electricity. 

riers of different velocities and different sources. The presence 
of both radium and thorium emanations was shown, and the 
specific velocities of the positive carriers of radium emanation 

were found to be in region between 25 and tt-t^t — ' / , 

6 40,000 sec./ cm.' 

while that of thorium emanation lay in the region between 15 

n cm / volt mi , , .. . 1 _ 

and 0*2 ■ — •' / . Ine negatively charged carriers could not 

sec./ cm. iD j n 

be collected in sufficient quantity to give a quantitative measure- 
ment. From the specific number of the carriers found in the 
atmosphere the ionization caused by them was computed and 
found to be only a small portion of the total ionization of the 
atmosphere. Most of the authorities agree in concluding that 
the radio-activity of the atmosphere can play only 'a minor role 
in its ionization, and that the ionization must be referred to 
other causes. My own experience at sea led me to the same 
conclusion, for there no measurable amount of induced radio- 
activity could be collected, while the ionization was as great 
as on land if not greater. Eve, on the other hand, reports that 
he has found the amount of emanation in the air over the 
Atlantic Ocean to be as great as over the land. Freshly fallen 
rain and snow on land ordinarily are very noticeably radio-active, 
but 1 could find no trace of radio-activity in rain water collected 
at sea. Observations on a mountain peak in South G-ermany 
at a height of 2954 meters at intervals throughout a year 
showed no trace of radio-activity in the rain water collected there, 
while it was present in that collected on the plains below, show- 
ing that the rain must pick up its radio-activity in the lower 
layers of the atmosphere. Observations in balloons have shown 
that the ionization increases with the altitude, proving that the 
ionization is not proportional to the emanation. It is a well- 
known fact that the ultra violet rays, the Rontgen, cathode and 
Becquerel rays are all capable of causing ionization, and it seems 
very plausible that coming from the sun they may produce a 
high degree of ionization in the upper regions of the atmos- 
phere,, and, to a less degree, in the lower regions, while in the 
lower regions the radio-active substances may add a certain 
portion to the ionization. In this connection the investigation 
of the problem of the spontaneous ionization of air in closed 
vessels may be cited. Campbell and Wood, at the Cavendish 
Laboratory, have made a long study of the subject and found 
that air enclosed in a vessel and left undisturbed is constantly 
being ionized and that the rate of ionization undergoes a peri- 
odic diurnal variation with pretty well-defined maxima and 
minima. Screening the vessel with masses of lead reduces the 



Dike — Recent Observations in Atmospheric Electricity. 209 

rate of ionization though it does not entirely stop it, the 
inference being that the ionization is caused by a penetrating 
radiation from without, which can be partially screened off by 
the lead. W. W. Strong, working at the John Hopkins Physical 
Laboratory, has investigated the same problem, and by screening 
the vessel on all sides except the top and noting the rate of 
ionization, arrived at the conclusion that part of the radiation 
comes from above. Observations in a cave gave a lower rate 
of ionization than in the open, a fact pointing in the same 
direction. 

There is a great field for investigation along all these lines 
and, as I have said, one that is almost entirely neglected in this 
country. A well-equipped observatory with opportunity for 
research work and the development of improved apparatus 
could add largely to our knowledge of the electrical condition 
of the atmosphere and aid in solving some of the puzzling 
problems, while a single observer here and there can accom- 
plish comparatively little. The need at present is for investi- 
gation, and persistent sustained attacks on single problems 
rather than extended field work, though this is of value when 
the atmospheric conditions encountered are such as can not be 
attained at the existing observatories, as on mountain peaks, at 
sea or in caves. 

Department Terrestrial Magnetism, 
Carnegie Institution of Washington. 



210 Kraus and Cook — Iodyrite from Tonopah. 

Art. XIII. — Iodyrite from Tonopah, Nevada, and Broken 
mil, New South Wales / by E. II. Kraus and C. W. Cook. 

In November last the Mineralogical Laboratory of the Uni- 
versity of Michigan received a consignment of minerals from 
the Foote Mineral Company of Philadelphia, Pa., which con- 
tained ten selected crystals of iodyrite from Tonopah, Nevada, 
a new locality for this mineral. Our attention was called to 
this fact by Mr. W. M. Foote, manager of the company. On 
examination, several of these crystals showed that they pos- 
sess a pronounced hemimorphic development, and also several 
forms which had not as yet been observed on iodyrite. We 
immediately informed Mr. Foote of these facts and he most 
cordially placed a very liberal quantity of selected material at 
our command. He subsequently sent us a specimen from 
Broken Hill, New South Wales, which contained crystals of a 
most interesting character. 

Through the courtesy of Director H. C. Bumpus and 
Curator L. P. Gratacap of the American Museum of Natural 
History of New York City, we w T ere able to examine speci- 
mens 2609 and 2610 of their mineral collection. The first of 
these is from Broken Hill, the second from Chile. 

The various contributions on iodyrite are rather fragment- 
ary and also widely separated in time. It was, therefore, 
deemed advisable to preface the results of our study with a 
brief survey of the work already done on this mineral. 

Historical. 

In 1825 Vauquelin* pointed out that certain silver ores from 
Mexico contained the element iodine, probably in the form of 
the iodide. This was the first record of the occurrence of the 
element iodine in the mineral kingdom, it having been previ- 
ously observed only in plants and animal remains. Cantuf 
had, however, showed the presence of iodine in traces in cer- 
tain mineral waters from Asti, Italy. That silver iodide occurs 
in nature was first definitely pointed out by Domeyko,J who 
in 1844 analyzed material from Chanarcillo, Chile. As to the 
crystallographic development of this material the only state- 
ment made by Domeyko is that rhombohedral-like forms were 
occasionally observed and also that the structure was more or 
less lamellar. Domeyko did not make a complete analysis but 
showed conclusively that the amount of silver in the natural 

* Annales de Chimie et de Physique, xxix, 99, 1825. 
f Annales des Mines (4), vi, 158, 1844. 
% Cited by Vauqiielin, loc. cit. 



Kraus and Cook — Iodyrite from Tonojpah. 211 

compound agreed very closely with that of the artificial prod- 
uct. Crystals from this same locality, which were in the 
mineral collections of the Ecole de Mines, Paris, were meas- 
ured by Des Cloizeaux* in 1854. Des Cloizeaux assigned the 
mineral to the hexagonal system and calculated the axial ratio 
to be a: c = l: 0*81438. The nine forms observed by him are 
given in the tabulation on page 21 Y. Although it is not clear 
that Des Cloizeaux observed hemimorphism, it is, neverthe- 
less, of interest to note that he pointed out the very close simi- 
larity of the angles of the natural iodide of silver and the 
hemimorphic mineral greenockite. 

In 1851 Danaf suggested the name iodyrite for this mineral, 
although Haidinger^: had previously (1845) called it iodite. 
The name iodyrite has become international in spite of the 
fact that Leymerie§ introduced iodargyrite in 1859. 

The next contribution to our knowledge of this compound 
was made in 18Y9 by Zepharovich,|| who carried out a rather 
extensive examination of artificial crystals, w T hich had been 
prepared by Belohoubek. In all Zepharovich observed twelve 
forms, seven of which had, however, been previously noted 
by Des Cloizeaux on natural crystals ; see the tabulation on 
page 21 Y. It is of some interest to indicate in this connection 
that although the hexagonal prism of the first order is, accord- 
ing to Des Cloizeaux, a common form on iodyrite, it was, 
nevertheless, not observed by Zepharovich on artificial crystals. 
Zepharovich established a new axial ratio which differed con- 
siderably from the one obtained by Des Cloizeaux. Accord- 
ing to Zepharovich the ratio is as follows: a:c=l: 0*8196. 
These values have been generally adopted. T The crystals 
examined by Zepharovich possessed a pronounced hemimorphic 
development. 

In 1881 Seligmann** examined crystals of iodyrite from 
Dernbach, Nassau, and observed seven forms ; see page 21Y. 
The development of these crystals was such as to point to 
hemimorphism in that £[2021 1, 6>jl011^, and ^j4041[ were 
observed above the prism of the first order while a' j 2021} was 
noted below. Seligmann also examined crystals from Chile 
and on these noted four forms which had not been observed 
on either natural or artificial crystals of silver iodide ; see page 

* Armales de Chimie et de Physique, xl, 85, 1854. 

f System of Mineralogy, 4th edition, 95, 1854. 

X Chester, A Dictionary of the Names of Minerals, 189G, 134. 

§ Chester, loc. cit. 

|| Zeitschr. Kryst., iv, 119, 1879. 

^[Dana, System of Mineralogy, 6th edition, 1892, 160; Groth, Tabella- 
rische Uebersicht der Mineralien, 4th Auflage, 1898, 50 ; Naumann-Zirkel, 
Elemente der Mineralogie, 14te Anflage, 1901, 505. 

** Zeitschr. Kryst. vi, 229, 1881. 



212 Kraus and Cook — Iodyrite from Tonopah. 



217. These crystals were apparently holohedral in their devel- 
opment. 

The work of Seligmann was followed in 1885 by a descrip- 
tion of crystals of iodyrite from Lake Valley, New Mexico, 
by Genth and vom Rath.* Only three forms were noted by 
them, but they describe an interesting type of twin crystal, 
the individuals of which possess a rhombohedral development. 
A plane parallel to a face of e\ 3034 \ acts as the twinning 
plane. A twin of this character is reproduced by Dana.f 

The only other work on iodyrite up to the present time, 
which has come to our knowledge, is that by Spencer J in 1901. 
Spencer describes crystals from Broken Hill, New South 
Wales. The simple crystals were tabular or short prismatic 
and showed practically the same forms as observed by Selig- 
mann on crystals from Chile ; see page 217. Spencer, however, 
noted extremely interesting twins possessing a pseudo-cubical 
development. These twins are interpreted by him as being 
composed of four individuals of a trigonal character. The 
twinning law is the same as that observed by Genth and vom 
Rath, namely, 6J3034J. Repeated attempts were made by 
Spencer to obtain etch figures but without success. 

Iodyrite from, Broken Hill, New South Wales. 

The crystals examined from this locality occur on a speci- 
men of limonite and psilomelane, which was placed at our dis- 
posal by the Foote Mineral Company of 
Philadelphia, Pa. The crystals are very 
small, being from 1 to 2 mm in length and 
in all cases show an apparently holohedral 
development. The color is a rather bright 
lemon yellow. The faces are not nearly 
as brilliant as those possessed by the crys- 
tals from Tonopah, Nevada, which will be 
described later. Although it was necessary 
in making our readings to rely almost 
entirely upon the maximum shimmer or 
luster, we, nevertheless, feel that consider- 
able confidence may be placed in the 
results thus obtained. The following forms, _ 
ojOOOl}, c'jOOOl}, m{1010\, « 1 1120 1, ^j 1011}, and u'\0£L\, 
were definitely determined. Another form of a pyramidal 
character was also noted, but our readings in this case were 
somewhat unreliable so that we do not feel warranted in mak- 
ing a statement as to its probable indices. The crystals possess 

* Zeitschr. Kryst., x, 473, 1885. 

f System of Mineralogy, 6th edition, 1892, 160. 

% Zeitschr. Kryst., xxxv, 460, 1901. 




Kraus and Cook — Iodyrite from Tonopah. 213 

a prismatic habit, a {1120}, being by far the predominating 
form. The prism m{1010} occurs as very narrow faces trun- 
cating the edges of ${1120}. This is the first time that the 
simultaneous occurrence of the prisms of the two orders on 
a single individual of either the artificial or natural compound 
has been recorded. Furthermore, the prism, a {1120}, had not 
previously been observed on natural crystals. The general 
type of development, possessed by these crystals, is clearly 
shown by figure 1. 

The averages of several of the more important readings 
used for the identification of the forms given above are as 
follows : 

Observed Calculated 

c : u = (0001) : (4041) 1h° 21' 75° 13' 

c : m = (0001) : (1010) 90° 4' 90° 00' 

m : a = (1010) : (1120) 30° 2' 30° 00' 

Through the courtesy of the American Museum of Natural 
History of New York, we have been able to examine speci- 
men 2609, which shows some excellently developed crystals. 
Our examination of this interesting specimen was, however, 
confined to the use of the hand lense, but it is evident that 
the crystals are undoubtedly twinned according to the law 
noted above. Not being able to carry out goniometric meas- 
urements on any of the individuals on this specimen, we are 
unable to state definitely what forms occur upon them. 

Iodyrite from Tonopah, Nevada. 

The material examined from Tonopah, Nevada, consisted of 
several thousand isolated crystals and two matrix specimens 
furnished by the Foote Mineral Company. From a study of 
the matrix specimens, composed almost entirely of quartz, on 
which the iodyrite occurs as isolated crystals or in crystalline 
crusts, it is clearly evident that the mineral is of secondary 
origin. 

The crystals vary considerably in color and size. Those 
studied more closely varied from a bright lemon-yellow through 
the darker shades of yellow to a yellow-green. They measured 
from 1 to 5 mm in length. The crystals possess a greasy ada- 
mantine luster and occur as simple crystals, parallel groupings, 
twins, and rosette clusters. In all nine types of development 
were observed, four of which appear to be characteristic of 
simple crystals. 

Type 1. Crystals of this type possess a pronounced hemi- 
morphic development. They show the following forms : 
c{0001} 5 m{10l0}, ^2021}, ^{2021}, £{7073}, w{9092}, 

Am. Jour. Sci.— Fourth Series, Vol. XXVII, No. 159.— March, 1909. 
15 



214 Kraus mid Cook — Iodyrite from Tonopah. 



w'{ 9092 }, y{9091}, and 2/'{9091[. Of these forms m|1010f 
predominates. The lower pyramid, ^ v 1 2021 } , is also rather 
large. The upper basal pinacoid, cJOOOIJ, is usually next in 
size. The other forms generally occur as rather narrow faces. 
The relative sizes of the faces of the various forms are clearly 
shown in figure 2. Of the forms observed on crystals of _this 
type,_it is well to point out that *|7073}, w{9092}, w'{9092}, 
2/{ 9091 } ? and y'{9091[, are all new. The close agreement 
between the calculated and observed values for these new 
forms makes us feel confident that they are to be considered 
as established. 

Type 2. A pyramidal habit is a striking characteristic of 
crystals of this type. The following forms were noted on 



Figs. 2-5. 




them : cjOOOl 



c/5 



15.0.15.8 



, c'jooo: 

Of these forms s' 



* -J 2021 



2021 L and 



15.0.15.8 £ was observed for 



the first time. The general character of crystals of this type 
is shown in figure 3, which consists of a combination of the 
forms just enumerated. Hemimorphism is very pronounced 
on these crystals. 

In several instances on crystals of types 1 and 2 the 
additional new forms were noted: r\70 f t4:\, a? {7071}, and 
2 {33.0.33.2}. These forms are usually to be observed as very 
narrow faces, and as they do not occur often, we did not deem 
it necessary to figure them. 

Type 3. Crystals of this type of development are not 
very common. They possess a tabular habit and are appar- 



Kraus and Cook — Ioclyrite from Tonopah. 215 

ently holohedral. The forms observed are c \ 0001 } , c'\ 0001 \ , 
raj 1010 1, ^j 2021 1, and ^{2021}. Figure 4 shows these forms 
in combination. 

Type 4. Due to the predominance of two parallel faces of 
the prism, m{1010J, crystals of this type are also tabular. 
They are, however, hemimorphic, as is clearly shown in figure 
5. The figure shows a c_ombinatio_n of the following forms : 
c{0001J, m{1010J, t{7073\, •{2021}, and i'{2021\. These 
crystals are by no means as common as those of types 1 and 2. 

Type 5. Parallel grouping plays an important part in the 
formation of crystals of this group. This is shown in figure 6. 
These crystals are made up of several simple individuals which 
possess, in general, a habit similar to that which is so charac- 
teristic of crystals of type 2. Although they show hemimor- 
phism, they are, nevertheless, more or less barrel-shaped, 
resembling to a large extent the development so often observed 
on crystals of corundum. The observed forms are cjOOOl}, 
mjlOlOJ, i{2021\, and i'{2021\. These crystals are rather 
common and are sometimes as much as 5 mm in length. 

Type 6. Crystals belonging to this group are more or less 
carrot- or top-shaped. This is illustrated by figure 7. As is shown 
by the illustration, these crystals are made np of a number of 
individuals possessing a tabular development similar to that of 
type 3 (figure 4). The crystals show the upper base c|000_l( 
rather large and are terminated below by the pyramid i' \ 2021 \ , 
thus giving the groups a decided hemimorphic development. 
In many cases these crystals show a skeletal development in 
that they are more or less hollow. In figure 7 we show the 
general character of these crystals and also indicate the skeletal 
development by figuring a depression in the upper base. In 
some cases this depression does not extend far into the crystal, 
while in other instances it extends almost the entire length of 
the same. The bottom of this depression, which is usually 
irregular in outline, but sometimes may be hexagonal, is a 
plane surface, which from the reflections obtained on the goni- 
ometer we know to be parallel with base CJ0001J. The sides 
of the depression are usually characterized by parallel and 
horizontal corrugations. These can be seen best in the crystal 
fragments. The forms _obseiwecl on crystals of this type are 
c{000r;, m-il010;, *{2021} 1 and * / J2021{. 

Type 7. Here belong contact twins with the pyramid 
e {3031 \ acting as the twinning plane. The two individuals 
are about equally developed and possess a tabular habit, as 
shown in figure 8. This tabular habit is due to the predomi- 
nance of two parallel faces of the prism m\ 1010 \ . These twins 
are rather common. 



216 Kraus and Cook — Iodyrite from Tonopah. 

Type 8. Figure 9 illustrates clearly the characteristic devel- 
opment of twins belonging to this group. Here one of the 
individuals is fully developed and generally somewhat tabular 
in habit. Hemimorphism is pronounced. Attached to the 
side of the fully developed crystal is the second smaller indi- 
vidual and in the position demanded by the common twinning 
law e {3034 1 . The second individual is also more or less tab- 

Figs. 6-9. 




C'^^^^^J 


/ ' ] 


J^ 


?n 


m 


\ i' 


— i 


\l m 


— ^H-- 































ular. Twins of this character were not as frequently encoun- 
tered as those of type 7. 

Type 9. Quite a number of contact twins, the individuals of 
which possess a rhombohedral development, were observed. 
They resemble very closely the twins from Lake Yalley, New 
Mexico, which were described in 1885 by Genthand vom Rath. 
The figure given by them is reproduced in Dana's System of 
Mineralogy.* 

In the writings of the earlier contributors to our knowledge 
of the crystallography of iodyrite much confusion exists as to 
the occurrence of the pyramidal forms about the upper and 
* 6th. edition, 1892, 160. 



Kraus and Cook — Iodyrite from Tonopah. 



217 



lower poles. Especially Zepharovich, in reporting upon the 
forms observed on artificial crystals with a decided hemimorphic 
development, is not at all clear in his statements as to the occur- 
rence of certain pyramidal forms. In the text these forms are 
indicated as occurring only below, while in the table of angles 
they are given as upper forms. It is also rather common for 
crystallographers, when describing the forms on hemimorphic 
crystals, to designate the upper and lower forms possessing the 
same indices by different letters. This is somewhat clearer 
but still confusing. 







Natural 


u 


Artificial 


Chile 


Dern- 
bach 


Lake 

Valley 


Broken Hill 


Tonopah 


-4-3 
-t-3 

<x> 


Zepharo- 
vich 


Des Cloi- 
zeaux 


Seligmann 


Genth 
&Eath 


Spencer 


Kraus and Cook 


C 


{ 0001 J 


{0001} 


{0001} 


10001} 


{0001} 


{0001} 


{0001} 


{0001} 


c' 


. . 


{0001}* 


{0001} 


{0001} 


. . 


{0001} 


{0001} 


{0001} 


m 





{1010} 


1010 


{1010} 




{1010} 


{1010} 


{1010} 


a 


{1120} 

















{1120} 





P 





{10.12} 




















r 


{1012} 


{1012}* 






















{1011} 








{10.11} 


{1011} 










o' 


{1011} 














.. 




i 


{2021} 


{2021} 


{20211 


{2021} 


{2021} 


{2021 





{2021} 


i' 


{2021} 


{2021}* 


{2021} 


{2021} 




{2021 




{2021} 


V 


{2023} 


. _ _ . 










. 








f 







{3031} 








{3031} 







f 


. . _ . 





{3031} 














. .. _ . 


g 








{3032} 








|3032} 







g\ 








{3032} 







t ' 


. 





e 


' {3034} 























u 


{4041} 


{4041} 




{4041} 








{4041 


.... 


v! 





{4041}* 













{4041 





r 
7T 


{4045} 


_ _ _ _ 











..-■- 







P 


{9.9.18.20} 





















r 






















{7074} 


t 






















{7073} 


X 






















{7071} 


y 























{9091} 


y' 








_ .. _ _ 














{9091} 


10 























{9092} 


10 ' 





. _ . _ 





_ _ . . 











{9092} 


s' 


. . _ - 
















. 


{15.0.15.8} 


z 





















{33.0.33.2} 



* Inasmuch as Des Cloizeaux makes no definite statement as to the develop- 
ment of these crystals, we have considered them as being apparently holohedral. 



218 Kraus and Cook — Tody rite from Tonopah. 

In order to avoid all confusion we have divided the basal 
pinacoid and the various bipyramids into upper and lower 
forms, following the practice of the more recent texts on crystal- 
lography. Certain letters are then assigned to the various 
bipyramids and the basal pinacoid, which, when unprimed, 
indicate upper forms, when primed, lower forms; thus, o J1011 J, 
upper; VjlOHJ, lower. We believe that this system of 
nomenclature offers decided advantages over what has been 
heretofore in use for hemimorphic crystals, in that it does away 
with all confusion. 

In accordance with the plan outlined in the preceding para- 
graph we have prepared a general summary of all forms thus 
far reported as occurring on either natural or artificial silver 
iodide. 

As previously stated, p. 211, the axial ratio generally adopted 
for iodyrite is the one which was established by Zepharovich* 
for artificial crystals. Inasmuch as it is always highly desirable 
to have ratios based upon reliable observations on the natural 
compound, we have taken a very large number of readings 
with the view of establishing a new axial ratio. The values 
thus obtained differ but slightly from those given by Zepharo- 
vich. In order to show the close agreement of the two deter- 
minations we place them below and at the same time also add 
the values obtained by Des Cloizeauxf on natural crystals in 
1854. 

a : c 

Des Cloizeaux 1 : 0-81438 

Zepharovich _ 1 : 0-81960 

Kraus and Cook. _. ___ 1 : 0-82040. 

In all fifteen crystals were measured. The observed and 
calculated angles based upon the axial ratio, given by us, show 
in nearly all cases very close agreement. 

Observed Calculated 

c : "i — (0001) : (2021) 62° 10' 30"* 

m : m = (1010) : (0110) 59° 59' 60° 00' 

c : r = (0001) : (7074) 58° 47' 58° 50' 

c' : s' = (0001) : (15.0.15.8) 60° 32' 60° 47' 

c : t = (0001) : (7073) ' 65° 33' 65° 39' 

e : w = (0001) : (9092) 76° 46' 76° 48' 

c : x = (0001) : (7071) 81° 22' 81° 25' 

c : y = (0001) : (9091) 83° 13' 83° 19' 

c : z = (0001) : (33._0.33. 2) 86° 20' 86° 21' 

i : i = (2021) : (0221) 52° 30' 52° 29' 

m : i = (1010) : (0221) 63° 43' 63° 45' 

* Loc. cit. f Loc. cit. 



Kraus and Cook — Iodyrite from Tonopah. 219 

Etch Figures and Class of Symmetry 

Iodyrite is generally* placed in the class of symmetry which 
is designated by Grothf as the dihexagonal pyramidal class. 
This is based entirely upon the geometrical development of the 
crystals, which, however, might allow of such interpretations 
as to permit assigning the compound to at least three other 
classes, since no faces of the most general form have as yet 
been observed, Hence, it is necessary to rely upon etch figures 
in order to determine with definiteness the symmetry of 
iodyrite. 

Spencer \ indicates that repeated attempts were made to 
obtain etch figures upon the crystals which he studied from 

Fig. 10. 




(p//o) 



(//oo) 



(/OTOJ 







(0//OJ 







{/ / oo) 



yo/o) 



Broken Hill, New South Wales. These attempts were unsuc- 
cessful. Spencer, unfortunately, does not mention the solvents 
which he employed. 

A solution of potassium iodide is considered by Roscoe and 
Schorlemmer§ as a good solvent for silver iodide. A solution 
of N/2 potassium iodide r was used by us for the production of 
etch figures on carefully selected crystals of iodyrite from 
Tonopah. Although the solution etches the crystals, we were 
unable after repeated attempts to obtain figures possessing a 
definite outline. 

According to Gmelin|| silver iodide is soluble in a solution 
of mercuric nitrate and in a concentrated solution of either 

*See Dana and Naumann-Zirkel, loc. cit. ; also Bauer, Lehrtrach der Miner- 
alogie, 2te Auflage, 1904, 439 ; Klockmann, Lehrbuch der Mineralogie, 4te 
Auflage, 1907, 403 : Groth, Tabellarische Uebersicht der Mineralien, 4te 
Auflage, 1898, 50. 

f Physikalische Krystallographie, 4te Auflage, 1904, 496. 

% Loc. cit. 

§ Treatise on Chemistry, 1907, ii, 468. 

f Handbook of Chemistry, 1852, vi, 157. 



220 Kraus and Cook — Iodyrite from Tonopah. 

potassium or sodium chloride. It was, therefore, decided to 
use a cold concentrated solution of sodium chloride. Immersion 
in such a solution for fifteen seconds produced excellent micro- 
scopic etch figures of an hexagonal outline on the base and 
located as shown in figure 10. Figures of the general form and 
position, as illustrated in the diagram, were produced on the 
prism faces after a treatment of about forty seconds. The posi- 
tion and outline of these figures show conclusively that crystals 
of iodyrite possess six vertical planes of symmetry and one polar 
axis of hexagonal symmetry. These elements show that the 
classification which has been followed, as indicated above, is 
to be considered as the correct one. 

Chemical Analysis. 

For the chemical analyses somewhat over a gram of clear, 
semi-transparent crystal fragments of a pale yellow color was 
used. This material was from Tonopah, Nevada, and was 
placed at our disposal by the Foote Mineral Company, to 
whom we desire at this time to express our thanks for the cor- 
dial manner in which assistance was rendered us during this 
investigation. 

As to the analysis only the method of decomposition need 
be given in detail. This consists of placing a small sample 
of the finely powdered homogeneous material in an evaporat- 
ing dish and covering the same with water acidified with one 
or two drops of concentrated sulphuric acid. A small piece 
of chemically pure zinc is then suspended in the solution so as 
to just come into contact with the powdered mineral. Decom- 
position is generally complete after about twenty- four 
hours. The hydriodic acid which is liberated passes into 
solution, while the silver collects in the dish in the finely divided 
metallic state. In this way any loss of hydriodic acid by vola- 
tilization was avoided. For the sake of convenience the zinc 
was attached to a piece of platinum wire, so that after the decom- 
position was complete, it could be readily removed and any 
adhering silver washed off. In order to remove any traces of 
silver which could not be washed off, the zinc was immersed for 
an instant in nitric acid. After filtration the metallic silver was 
dissolved in nitric acid and then precipitated and weighed as 
the chloride. Silver nitrate was added to the filtrate and the 
iodine precipitated and weighed as the iodide. 

Two analyses were made which show very close agreement 
and indicate that the composition of iodyrite may be expressed 
by the formula Agl. 

In 1854 J. Lawrence Smith* made two analyses of iodyrite . 

* Dana, System of Mineralogy, 4th edition, 1855, 506. 



Kraits and Cook — Iodyrite from Tonopah. 221 

The localities from which the material was obtained are not 
indicated by Smith. We give below the average results of 
Smith's analyses. Smith reports that aside from silver and 
iodine, he found traces of copper and chlorine. Repeated 
tests were made by us for bromine and chlorine, as also for the 
metals lead and mercury, which would most likely replace the 
silver. But in no case did we find the slightest trace of any of 
them. 

The results of the analyses are as follows : 





Kraus and Cook 


Smith 

46-45$ 
53-02 ' 

99-47 


Theoretical 


Ag_.._ 
I .__. 


I 

.. 45*87^ 
._ 53-92 

.: 99-79 


II 

46-04$ 
54-09 


Average 

45-95$ 
54-01 

99-96 


45-97$ 
54-03 


Total . 


100-13 


100-00 



We are indebted to Professor E. D. Campbell, Director of 
the Chemical Laboratory of this University, for suggestions 
relative to methods of decomposition and analysis. , 

Specific Gravity. 

Considerable variation exists in the values given for the 
specific gravity of iodyrite or the artificial compound. This 
variation is clearly shown by the following tabulation : 

Iodyrite. 

Bauer* _. _. 5 5 — 5-7 

Damourf 5-677 

Dana* 5-6 — 5-7 

Domeyko* 5*504 

Klockmann* ._ 5*707 

Naumarm-Zirkel* 5 - 6 — 5*7 

VomRath* 5-609 

Artificial Silver Iodide. 

Damourf. _ 5*669 

Deville 5-540 

Since the material which we used for the chemical analysis 
proved to be exceptionally pure, we feel that the values we 
obtained for the specific gravity are very reliable. Two 
determinations were made on material which differed slightly 
in color. There is also a slight difference to be noted in the 

* Loc. cit. 

f Eammelsberg, Handbuch der Krystallographisch-Physikalische Chemie, 
I, 303, 1881. 



222 Kraus and Cook — Iodyrite from Tonopah. 

results. The determinations were made with the common 
specific gravity flask at room temperature. 
The results are : 

Dark variety ..__ 5*519 

Light variety 5*504 

Average 5*5115 

These values are much lower than those given by Damour, 
Dana, vom Rath, Klockmann, and Naumann-Zirkel, but agree 
fairly well with the determinations of Domeyko on natural 
crystals and with the value given by Deville for the artificial 
compound. Our results also fall within the limits set by 
Bauer. 

Mineralogical Laboratory, 

University of Michigan, 
December 22, 1908. 



H. S. Uhler — Deviation of Pays by Prisms. 223 



Art. XIV. — On the Deviation of Rays by Prisms; by 

H. S. Uhler. 

While making certain deductions from the formulae for the 
directions of rays passing obliquely through prisms, as given 
in standard text-books,* the writer found that conclusions 
which are at variance with physical facts were unavoidable. 
An investigation of the source of this difficulty showed that 
the primary error consisted in tacitly changing the definition 
of deviation in passing from considerations which refer to 
principal sections of prisms to cases where the rays traverse 
the prisms outside of these sections. If it were never neces- 
sary to combine the formulae for rays in principal sections 
with formulae for rays not in such planes, the error would 
only amount to inconsistency of notation and hence it would 
hardly deserve formal notice. As a matter of fact, how- 
ever, it is sometimes necessary to combine formulae for the 
two cases and; when this is done, false results follow. There- 
fore, it may not be superfluous to give a brief discussion of 
the fundamental formula for ra}^s not in principal sections in 
such a manner as to retain the definition of deviation in the 
form which is usually, if not universally, given to it when 
principal sections alone are concerned. 

This definition may be concisely stated as follows : The 
deviation of a ray is the angle through which the unlocalized 
vector representing the initial direction of the ray must be 
turned in order to make it coincide in sense and direction 
with the like vector ivhich symbolizes the filial direction of the 
ray. This sentence expresses precisely what is meant by the 
deviation of light when making observations with an ordinary 
spectrometer or even when considering the passage of plane 
waves through a plane-parallel layer of some transparent me- 
dium. On the other hand, the above definition makes the angle 
of deviation supplementary to the deviation involved in the 
current formula for obliquely transmitted rays. 

If D and D' denote respectively the deviations, produced by 
a single prism of greater index of refraction than the surround- 
ing medium, of a ray not in a principal section- and of the 
projection of the same ray on a principal section, and if \ 
symbolizes the angle which the incident (or emergent) ray 
makes with the plane of such a section, then the mutual 
dependence of these quantities upon one another is given by 

siniD=sin JD' eos^ (1) 

* Czapski, Heath, Kayser, Winkelmann, etc. 



224 H. S. Uhler — Deviation of Rays hy Prisms. 

This formula may be obtained in two different ways. Either 
the classic relation " cos^D^cos-JD'cos i" may be first 
demonstrated in the customary manner and then D aud D' 
may be replaced respectively by tt— J) and it — D', or else an 
independent proof may be given. The latter plan seems to be 
preferable because it is more direct, and also since it does not 
involve a knowledge of a formula of spherical trigonometry, 
whereas the usual proof does. In other words, the direct 
method is even more elementary than the indirect. Con- 
sequently an independent proof of equation (1) will now be 
presented. 

Fig. 1. 




Figure 1 may serve to define the positive directions of the 
various angles as well as the symbols for the angles themselves. 
The use or omission of accents on small letters shall serve to 
differentiate respectively symbols which refer to angles in the 
principal section from those which denote angles lying outside 
of this plane. The path of the ray is AB'EF. The projection 
of the ray on the principal section is A'B'C'F'. Further, let 
D symbolize the deviation of the ray from its initial direction 
AB V to its final direction EF. 

Move the directed lines AB', A'B', EF and O'F', of figure 1, 
parallel to themselves until they assume respectively the posi- 



R. S. U filer — Deviation of Bays by Prisms. 225 

tions OB", OB 7// , OF" and OF'", that is, all starting from any 
common point O. Lay off OB /!r =OB ,,, =OF / '=OF ,/, =L. 
£B'"OB' , = i 1 , ^F"'OF # =» 4 , ^B'"OF"'=D' and ZB"OF"=D. 

Since, from the general theory of prisms, i 1 =i 49 a triangle 
B"GF" may be constructed in a plane containing B" and F" 
and perpendicular to OG-. Join B ;// to F //; . 

Now in the triangle B"GF", B^^OF'siniD' and, in 

the triangle B"OF", FF"=2L sin^D. Also GF'=L cos i„ 
hence 

sin -|D = sin ^D'cos i x . 

Furthermore, when cos ^<1, equation (1) shows that the devia- 
tion of a ray not in a principal section is less than the devia- 
tion of its projection on such a plane, whereas the text-books 
agree in writing " I3>D r . " 

We shall next consider what bearing formula (1) has on the 
question of minimum deviation. The well-known, general 
equations for refraction by prisms are 

sin £ =w sin z' ) -. ( *,=£, 

. - 1 • . 2 y and ■< . 2 . 3 

sln^ 4 =?^sm^ 3 j ( ^ 1 = ^ 4 

cos ij sin /3\=n cos i t sin /3' 2 

cos i t sin /3' 4 =n cos i 3 sin /3' 3 

*' = £'. + 0', 

where n denotes the ratio of the absolute index of refraction 
of the material of the prism to the absolute index of the sur- 
rounding medium. In general, we shall assume w> 1. By 
using the equations of the above list together with the relation 

cos fH\ cos )3' 8 = cos p\ cos /3' 4 ; 

which is a necessary condition that D / shall be either a maxi- 
mum or minimum, when i x is kept constant, we find /3' 2 = 
$' % =\a! and hence 

cos t, sin ^(D' + a') = ( + /V /7i 2 — sin"*,) sin \d. (2) 

D' denotes the possible stationary value of D'. 
Equation (2) is equivalent to the usual formula 

"cos i, sin $(D , a + a')=n cos i 2 sin |a'," 

since n sin * 2 =sin *, and ^ 2 >-Jtt. It should be observed that, in 
obtaining equation (2), use has to be made of the relation 
D'=fi\ + ft\ — a' and this shows that the various writers intend 
to employ the same definition of deviation as the one given 
above. 

d 2 D' 

If we actually test , a , under the specified conditions, we 



226 H. S. U liter — Deviation of Rays oy Prisms. 

find that D\ is a true minimum, as would be expected from 
our knowledge of the existence of a minimum when i 1 =0, 
that is, when the projection of the ray coincides with the ray 
itself and the latter lies in a principal section. Hence, we 
can conclude from equation (1), since sines affect both D and 
D', not only that D has a stationary value simultaneously with 
D / but further that D attains a minimum value D at the 
same time that D' acquires its minimum D' . 

Again, since by hypothesis n exceeds unity, 

cos i x 

> n. Hence equation (2) shows that when the deviation of 
the projection of a ray not in a principal section assumes a 
minimum value, this value is greater than the minimum of 
deviation for rays in a principal section. Now equation (1) 
gives D < D' , [i x ^ 0), and equation (2) implies D' > A , where 
A represents the minimum deviation for principal sections. 
Consequently, on the face of it, nothing can be concluded 
about the relative sizes of D and A - For example, how do 
we know that the deficit of D with respect to D f may not be 
numerically greater than the excess of D' over A so as to 
make the minima of deviation for rays not in principal sec- 
tions less than the minimum of deviation for rays in principal 
planes ? In other words, how can we logically deduce the 
usual and correct theorem that : " The deviation of a ray by a 
prism is least when the ray passes through the prism in a 
principal plane and when the angles of incidence and emerg- 
ence are equal " ?* 

The usual argument is to say that D r exceeds A in value, 
that the equation " cos JD= cos JD'cos i" shows D to be 
greater than D', that D passes through a minimum value 
simultaneously with D r and that therefore a fortiori D > A . 
This assumes that D has a minimum, whereas both the D and 
the I)' involved in the cosine formula pass through maxima 
simultaneously and the general theorem does not follow. 

Consequently we shall now outline the proof of a formula 
for D as a function of a\ i x and n in order to see explicitly 
what the properties of D are. This formula may be obtained 
as follows: First, expand sin %(T)\ + a') of equation (2) in 
terms of sines and cosines of the angles %D\ and \a' . Next, 
reduce all cosines to sines, except cos -J-a 7 , and, after proper 
transposal of terms, square the members of the resulting 
equation. A quadratic in sin JD ; results and this equation 
is then solved for sin -JD^, care being taken to retain only the 

*R. S. Heath, Geometrical Optics, p. 32, 1887; or Ozapski, Kayser, 
Wink elm arm, etc. 



: 



H. 8. TJhler — Deviation of Rays by Prisms. 227 

proper sign before the radical. In this manner we find, after 
substituting in formula (1), that 



[ + V(^ a_ sin2 *i) cos2 i a '~\~ 



[ + ^/{n 2 — sin 2 *J cos 2 ia' — (n* — l)]}sin p.' (3) 

Obviously, this equation could not have been obtained by 
substituting in " cos -|D — cos iD' cos i'\ and hence, if formula 
(3) is correct, the cosine equation must be erroneous. 
Again, formula (3) leads to 

d D o sin 2 i x sin \a c os 2 \a! 

zzz X 

di^ cos %D 

( 1 1 



+ A /(n 2 -sin 2 ^ 1 )cos 2 ^a'-(^ 2 -l) + ^/(n 2 -sin^'Jcos 2 -^' 



(4) 



Since D cannot exceed 7r, and since the first fraction within 
the braces of equation (4) is larger than the second, it follows 
that D decreases when i x diminishes in absolute value. There- 
fore, the minima D of D decrease as i x approaches zero in 
magnitude until, when ^=0, D =A - Thus we have shown 
formally that, although D is always less than D' , D is con- 
stantly greater than A . The general theorem of minima 
quoted above from Heath is therefore established. 

Equation (3) is interesting because the second radical indi- 
cates limitations upon a\ i x and n in order that D be real, that 
is, in order that a minimum of deviation may exist. For 
illustration, when a' and ware given 



i x <sin- 1 [± v /l-(^ 2 -l)tan 2 ^a'], 

n 2 -2 

which m turn requires cos a' > — . 

n 

Furthermore, combining . with equation (2) the condition 
that the second radical of formula (3) shall vanish, we find 
sin J-(D' -f a')~l, so that D\ and a' are then supplementary 
in value. Of course, the same restricting conditions can be 
obtained directly from equation (2) by observing that 
S ini(D'„ + a')<i: y 

Returning from this digression, attention may be called to 
the fact that when a=0 or when n=l formula (3), which is 
implicitly formula (1), gives D =0, whereas the relation 
" cos iD = cos -J-D 7 ,, cos i", combined with the auxiliary equa- 
tion (2), leads to D =±i 1 for both cases. For a plane-parallel 
layer of relative index n, or for a prism of finite angle but 
with the same refractive index as the surrounding medium, 



228 H. S. Uhler — Deviation of Hays hy Prisms. 

it is obvious that D cannot be a function of i x . Hence, the 
cosine formula again leads to false conclusions. 

In order to test the accuracy of the algebraic work which led 
to equation (3) as well as to get a concrete idea of the varia- 
tions of D and D\ with i x , the numbers in the following- 
table were worked out for the special case where a'— 60° and 
n=l m 65. The data in the second column were obtained from 
equation (2). The numbers in the third and fourth columns 
were found by substituting the corresponding values of D' 
from the second column in equation (1) and in " cos |-D — 
cos iD\ cos i" respectively. All the angles in the third 
column were also worked out from formula (3). The abso- 
lute agreement of all the values of the true D obtained by 
these two independent modes of calculation verifies equation 
(3). The last row of the table exhibits the superior limits of 
the angles for minimum deviation. The first and fifth col- 
umns taken together show the obliquity of the incident ray. 



ii 


Do 


Do 


"Do" 


|3'i=i(D' + a') 


0° 


51° 10' 37" 


A ; 51° 10' 37", 


51° 10' 37" 


55° 35' 18".5 


± 5° 


51° 34' 57" 


51° 22' 19" 


52° 28' 40" 


55° 47' 28"'5 


+ 10° 


52° 49' 54" 


51° 58' 7" 


56° 14' 6" 


56° 24' 57" 


± 15° 


55° 1' 46" 


53° 0' 17" 


62° 6' 37" 


57° 30' 53" 


± 20° 


58° 22' 54" 


54° 33' 21" 


69° 45' 36". 


59° 11' 27" 


± 25° 


63° 15' 56" 


56° 45' 45" 


79° 0' 19" 


61° 37' 58" 


± 30° 


70° 24' 58" 


59° 54' 29" 


89° 55' 13" 


65° 12' 29" 


± 35° 


81° 35' 0" 


64° 42' 35" 


103° 20' 34" 


70° 47' 30" 


± 40° 


105° 23' 7" 


75° 4' 45" 


124° 40' 17" 


82° 41' 34" 


40° 44' 5" 


120° 0' 0" 


82° 1' 28" 


135° 28' 21" 


90° 0' 0" 



The conclusion to be drawn from the preceding argument 
is that either the formula "cos -|D=cos -J-D' cos i" must be 
replaced by sin %D= sin -J-D' cos i 1 or that the cosine formula 
may be retained, but, when the latter alternative is followed, 
the various writers should state explicitly, and lay special 
Stress upon, the fact that the D and the D' of the cosine rela- 
tion are the supplements of the corresponding deviations as 
involved in the other equations of the subject. We think 
that, for sake of consistency and to avoid confusion of symbols, 
the formula sin %D= sin -JD' cos i t should be introduced in 
the text-books in place of the cosine equation. 

Sloane Physical Laboratory, Yale University, 
New Haven, Conn. 



W. G. llixter — Heat of Oxidation of Tin. 229 

Art. XV. — The Heat of Oxidation of Tin, and second paper 
on the Heat of Combination of Acidic Oxides with Sodium 
Oxides j by W. G. Mixter. 

[Contributions from the Sheffield Chemical Laboratory of Yale University.] 

Tin. 

The heat of oxidation of tin obtained by different investi- 
gators varies widely and the work was done before calorimetry 
was perfected by Tkomsen, Berthelot, and others. Moreover, 
the constants of oxidations are required in calculating the heat 
of formation of sodium stannate ; hence it seemed desirable 
to make new determinations. Depretz* found for the heat of 
oxidation of the stannic oxide 170000°. Dulong,f in 1838, 
determined the heat evolved by the union of one liter of oxygen 
with tin and obtained 6411 c , 6325 c and 6790 c ; mean 6509 c , 
or for 32 grams of oxygen 145600°. Andrews,;}; who first 
invented a calorimetric bomb, made in 1848 good determina- 
tions. He mixed tin with broken quartz in a copper bomb, 
ignited the metal with a milligram of phosphorus and found 
the gain in weight. His results for one gram of oxvgen were 
4235°, 4244 c and 4210 c .; mean 4230 c , or for 32 grams of oxygen 
135360 c at constant volume and 136000° at constant pressure. 
Likewise for the combustion of stannous oxide he obtained 
4353 c , 4328 c and 4364 c ; mean 4349°, or for 16 grams of oxygen 
695S4 C at constant volume and 69900 c at constant pressure. 
The writer has used essentially Andrews' method. 

Experiments were first made with tin from a sodium alloy. 
Dr. C. H. Mathewson, to whom the writer is indebted for 
much good material for use in investigations, made a consider- 
able quantity of an alloy having approximately the composition 
Xa 2 Sn. This was pulverized and dropped into absolute 
alcohol. After the reaction had moderated, water was added 
gradually and then the metal was subjected to boiling water 
for several hours and washed thoroughly. The powder was 
grey and when magnified appeared to be made up of minute 
leaves. The stannic oxide from 1-616 grams of the metal 
weighed 2*060 grams, equivalent to 1'623 of tin. It was found 
to contain 0*12 per cent of sodium. This amount of impurity 
affects the heat result but slightly. For the calorimetric 
experiments the tin powder was placed in a weighed silver 
foil tray which was supported in the middle of a 500 cc bomb 
in order that, as the tray melted, the hot powder would fall 
through the oxygen. The weight of oxygen taken up was 

*Landoldt-Bornstein, Physicalisch-Chemische Tabellen, refer to Ann. Ch. 
Phys., xxxvii, 180. This reference the writer has not been able to find. 
+ C. R., rii, 871. {Phil. Mag. (3), xxxii, 321. 

Am. Jour. Sci. — Fourth Series, Yol. XXYII, No. 159. — March, 1909. 
16 



230 IF. G. Mixter—Reat of Oxidation of Tin. 

derived from the weight of tin taken, of oxides formed, and 
of ferrous-ferric oxide from the iron used for ignition. The 
experimental data are as follows : 

1 2 

Tin 8*121 grams 11 -070 grams 

" equivalent to oxygen com- 
bined _ 7-122 " 9-364 " 

Oxygen combined 1*910 " 2*518 " 

Water equivalent of system 3631* " 3482* " 

Temperature interval .. _ 2*366° 3*239° 

Heat observed 8591° 11279° 

" of oxidation of iron —80° — 80 c 

S511 c 11191° 

For 1 gram of oxygen 4456 c 4444 c 



The combustions were evidently incomplete as indicated by 
the figures and by the fact that there was formed a considerable 
amount of a black substance along with the white stannic 
oxide. The results are 4 per cent higher than those obtained 
with tin foil in experiments 3 and 4. It might be surmised 
that tin separated from a sodium-tin alloy is an allotropic 
form and it may be when the separation is made at a low 
temperature, but the powder which had been heated to 100° 
is the ordinary modification, for, as shown later, it gives the 
same heat when burned with sodium peroxide as ordinary 
crystalline tin. The high result is evidently due to the forma- 
tion of a considerable amount of stannous oxide. But the 
heat effect of Sn-f-0 is less than that of SnO + O, and hence it 
appears probable that the two oxides formed in the com- 
bustion had combined with evolutions of heat. Moreover, 
the sesquioxide of tin is known and is dark colored. 

The next experiments were made with tin foil, which a 
qualitative analysis showed to be quite pure. It was also 
tested as follows : 2*911 grams were treated with nitric acid, 
the latter removed by evaporation, the residue digested with 
dilute nitric acid and washed on a filter and finally heated 
over a blast lamp until the weight was constant. The stannic 
oxide obtained weighed 3*700 grams, equivalent to 2 916 
grains of tin. (Sn=119.) For the combustions the foil was 
placed in loose rolls in the bomb, which was then filled with 
dry oxygen at a pressure of 12 atmospheres. The conditions 
were the same as in experiments 1 and 2 except that the foil 
exposed a larger surface to the oxygen. The stannic oxide 
resulting was in the form of a porous cake which was brown- 
ish on the surface but white in the interior. The entire cake 



W. G. Mixter — Heat of Oxidation of Tin. 231 

was a mass of elongated microscopic crystals. Professor W. 
E. Ford of the Mineralogical Laboratory has kindly examined 
the crystals and found them too small for measurement but 
having the optical properties of cassiterite. It should be 
stated that a small portion of the stannic oxide was deposited 
on the upper surface of the bomb in the form of white 
powder. After a combustion the oxide was transferred to a 
platinum dish and the water used in washing out the last 
portions was evaporated. The weight of the oxygen which 
combined the tin was found as already described. The follow- 
ing are the experiments : 

3 4 

Tin ._ 15*168 grams 14*055 grams 

" equivalent to oxygen com- 
bined „__ 15-042 " 13-919 " 

Oxygen combined 4*045 " 3'743 " 

Water equivalent of system 3548- " 3658* " 

Temperature interval". 4-893° " 4-410° " 

Heat observed 17360 C 16132 c 

" of oxidation of iron . 41 c 40 c 



17319 C 16092 



For 1 gram of oxygen 4282 c 4293 c 

The average of the results is 4288 c for one gram of oxygen 
combining with tin to form stannic oxide. For 32 grams it is 
137200 c at constant volume and 137800 c at constant pressure. 
It will be observed that the tin in the form of foil was almost 
completely oxidized. Andrews' result was 136000° at constant 
pressure. In his experiments there may have been formed 
some amorphous stannic oxide which would lower the heat 
effect. His result, however, agrees well with the writer's. 

Stannous Oxide. 

A number of preparations of stannous oxide were made by 
different methods but only two appeared to be good enough for 
the purpose. The first used was made by adding an excess of 
ammonia to a solution of stannous chloride and then heating the 
mixture several days on a steam bath. The dark crystalline mass 
was washed thoroughly and dried at 100° and then heated in a 
current of dry carbon dioxide to about 400° as long as water 
and ammonia came off. The product was almost black, showing 
under the microscope minute crystals but no amorphous powder. 
It contained considerable stannic oxide, a trace of ammonia, and 
0*09 per cent of water. Three determinations of the heat of 
oxidations gave respectively : 4668 c , 4667 c and 4632 c , a mean of 



232 W. G. Mixter—Heat of Oxidation of Tin. 

4656 c for one gram of oxygen taken up or 74500 c for sixteen 
grams. The result was thought to be high and therefore 
another lot of stannous chloride was prepared as follows : 
about 400 grams of pure stannous chloride were dissolved in 
four liters of hot water and a solution of pure sodium hydroxide 
was added in sufficient quanity to dissolve part of the stannous 
hydroxide formed. The mixture was kept hot until most of 
the white particles had disappeared. The product was washed 
by decantation in order to remove the smaller particles present. 
The stannous oxide was in the form of minute dark crystals. 
It was free from chlorine and sodium; dried at 100° it con- 
tained 0*41 per cent of water, which was determined as follows : 
a weighed portion was burned in dry oxygen in order to attain 
a high temperature and the water was absorbed in a chloride 
of calcium tube. The determination of stannous oxide was 
made thus : 1*1193 grams of substance were heated in air and 
oxygen. The stannic oxide formed weighed 1*2425 grams and 
the observed gain in weight was 0*123 gram. Adding to this 
last number 0*0046 gram of water present before burning 
we have 0*1276 gram of oxygen taken up, which is equivalent 
to 96*2 per cent of SnO. 

For the calorimetric work the stannous oxide, dried at 100°, 
was placed on a silver foil tray which was supported at the top 
of the lower silver cup or lining of a 500 cc bomb and then the 
cup and contents were counterpoised on a balance. After a com- 
bustion the cup and contents were heated to expel moisture, 
allowed to cool and then the increase in weight noted. At the 
time of a combustion the tray melted and the oxide fell to the 
bottom of the cup, where it was found in the form of a porous 
white crystalline cake. The experiments were as follows : 

8 9 10 

Substance taken 30*06 30*89 30*58 grams 

Amount + 0*962 = SnO__. 28*92 29*72 29*49 
SnO equivalent to oxygen 

consumed 28'74 29*74 29*4 

Increase in weight 3*340 3*455 3*415 

Water in substance taken _ 0*123 0-126 0*125 

Fe o 0, from iron —0*056 —0*056 —0*056 



Oxygen consumed 3*407 3*525 3*484 

Water equivalent of system 3422* 3430* 3476* 

Temperature interval 4*422° 4-541° 4*469 

Heat observed .. _ __. 15133 c 15576 c 15534 c 

" of oxidation of iron.. — 64 c — 64 c — 64 c 



15069 c 15512 c 15470 c 



W. G. Mixter—Heat of Oxidation of Tin. 233 

For 1 gram of oxygen con- 
sumed .___ 4423 c 4401° 4443° 

" 1 gram of SnO in sub- 
stance taken 521 c 522° 521 c 

The average is 4417° for one gram of oxygen and for 16 grams 
it is 70672° at constant pressure and 71000° at constant volume. 
The mean for one gram of SnO in the substance taken is 
521 c ; for 135 grams 70335° at constant volume. Evidently 
the stannous oxide was nearly all burned to stannic oxide. 
Andrews' result was 69000 at constant pressure. 

Combustion of Tin with Sodium Peroxide. 

The metal used in the next three experiments was from the 
same lot taken for experiments 1 and 2 and prepared from the 
sodium-tin alloy. Carbon was added to the mixture in experi- 
ment 10 to ensure a high temperature, but it was not required, 
as the other experiments prove. JSTo metallic tin was found in 
the residues of the following determinations : 

11 12 13 

Tin... _.. 3-000 5-000 8-000 grams 

Carbon 0-330 " 

Sodium peroxide 10- 10- 18* " 

Water equivalent of system 3060- 3012- 2987" " 

Temperature interval 2-283° 1*871° 3-086° " 

Heat observed . 6986° 5635° 9218° 

" of oxidation of carbon —3663° 

" " " " iron.. —30° —30° —40° 
" " oxygen set free or 

taken up +46° —48° -f 



! V 



3339° 5617° 9260° 

For 1 gram of tin 1113° 1115° 1157° 

As the tin from the sodium-tin alloy gave more heat when 
burned with exygen than that known to be the common modi- 
fication, tin turnings were made by fastening a cylinder of 
tin in a lathe so that it rattled. Small chips came off which 
were separated from the coarser process by sifting. The 
following are the experimental data : 

14 15 

Tin 8*241 grams 10*175 grams 

" not burned 0-243 ". 0-064 " 

" burned ._.„ 7*998 " 10-111 " 

Sodium peroxide 22* " 25* " 

Water equivalent of system. __ 3759* " 3816* " 

Temperature interval 2*383° 3*020° 



234 W. G. Mixter — Heat of Oxidation of Tin. 

Heat observed . _ 8957° 11525° 

" of oxidation of iron — 64 c — 64 c 

" " oxygen taken up — 90 c 



8893 c 11371° 

For 1 gram of tin .. 1112 c 1125 c 

The mean of the five results is 1124 ; the average for 119 
grams of tin is 133756°. Evidently the tin from the sodium- 
tin alloy is the common modification. 

Amorphous stannic oxide used in the next two experiments 
was made by heating metastannic acid over a ring burner until 
the weight was constant. Sulphur was used to give the tem- 
perature requisite and to reduce the sodium peroxide to oxide. 



Stannic oxide, amorphous. .. 

" " -unchanged. . 

" " combined _ . . 
Water equivalent of system. 
Temperature interval . 



Heat observed . 

" of oxidation of iron _ . 
" " " " sulphur. 

" " oxygen set free .... 



• 16 




17 


9-955 gi 


rams 


7-907 grams 


1-204 


u 


0-583 " 


8-701 


i( 


7-324 " 


2987- 


a 


2987' " 


2-500° 




2-337° 


7467 c 




6961 c 


— 48 c 




— 48° 


— 5271° 




— 527l c 


+ 75° 




+ 75 c 


2223 c 


1717 C 


255 c 




237 c 



For 1 gram of stannic oxide . 

The mean is 246 c and for 151 grams of amorphous stannic 
oxide it is 37100°. 

Summary of results. 

Sn + 2 = Sn0 2 (crystalline) + 137200° 

Sn (crystalline) + O = SnO(crystailine) + ... 66200 c 

SnO(crystalline) + O = SnO.. (crystalline) + __ 71000° 

Sn0 2 (amorphous) = Sn0 2 (crystalline) + . 1700° 

2Na 2 2 + Sn = Na 2 Sn0 3 + Na.O + -- 133800° 

2Na 2 + 20 = 2Na 2 + . 38800° 

Na 2 + Sn + 20 = Na 2 Sn0 3 + 172600° 

Sn + 2 = Sn0 2 (crystalline) + .'_ 137200° 

Na 2 4- Sn0 2 (crystalline) — Na 2 Sn0 3 + 35400° 

Na 2 + Sn0 2 (amorphous) = Na 2 Sn0 3 + 37100° 



Ford — Neptunite Crystals from San Benito Co., Cat. 235 

Art. XYI. — Neptunite Crystals from San Benito County, 
California ; by W. E. Foed. 

Historical. — In July, 1907, a new mineral, called benitoite, 
was described by Prof. G. D. Louderback in a bulletin of the 
Department of Geology of the University of California.* In 
the same article a short preliminary description of what was 
then thought to be another new species was given and to 
which the name carlosite was provisionally assigned. Mr. 
Lazard Calm, in the summer of 1907, first called the writer's 
attention to the fact that this mineral was identical with the 
rare species neptunite found previously only in the Juliane- 
haab district, Greenland. This identity of carlosite with nep- 
tunite was announced by Prof. Louderback at the meeting of 
the Geological Society at Albuquerque in December, 1907, 




m 



m\ 




Fig. 3. 






and a short note to the same effect was later published by W. 
E. Blasdale in Science for August 21st, 1908. But beyond 
this no description of the occurrence has so far been published. 
Through the courtesy of Mr. Calm and Mr. Milton G. Smith, 

*Vol. v, No. 9, 149, 1907. 



236 Ford — Neptunite Crystals from San Benito Co., Gal. 



Fig. 4. 



the writer has been able to examine a large number of the 
neptnnite crystals from this locality and it was thought that a 
brief description of them would be of interest. 

Occurrence. — The neptunite crystals occur associated with 
those of benitoite, both minerals being embedded in a white 
matrix of crystalline granular natrolite. The neptunite crys- 
tals that are now in the Brush Collection vary in size from 
slender prismatic crystals of only a few millimeters in length 
to those of 2-| cm in length and as much as 7 mm in thickness. In 
general the crystals are brilliant black in color, but wherever 

they have been fractured they show 
by internal reflections the red-brown 
color which is characteristic of the 
mineral when in thin section. 

Crystallograjphic. — The crystals 
from San Benito County are excep- 
tionally uniform in their habit, the 
following series of forms being found 
on almost every crystal studied : 
a(100Vm(110), 5(111)^(112), tf(lll), 
^r* (211), j? (311). The crystals are 
prismatic in habit, quite different 
from those first described by Flink* 
from Greenland, but more like a later 
type described by him'f and resem- 
bling still more a crystal figured by 
Wallenstrom4 The habit of devel- 
opment of the California crystals is 
represented in the figures which have 
been drawn from specimens in the 
Brush Collection. Figure 1 repre- 
sents the commonest type in which the 
forms present are : prism, m (110), 
prominent, a (100) in narrow trun- 
cations, the base, c (001) very small, 
as also 5(111) and * (112). The 
prominent terminal faces are the 
negative pyramids o (111), the new 
Figure 2 shows a slight modification 
in the development of these forms, and figure 4 shows the same 
forms, with the exception of g which failed, but in an unsym- 
metrical development which is unusual on the crystals. This 
figure was drawn by Mr. D. D. Irwin. Figure 4 is a doubly 
terminated crystal, like figure 2, half embedded in its matrix 
of white natrolite. The faces of the crystals studied were, as 
a rule, bright and gave excellent reflections on the goniometer 
with the exception of the new form g (211). In every crystal 

*Zeitschr. Kr., xxiii, 346, 1894. fMedd. om Gronl., xxiv, 120, 1902. 
{Geol. For. Forh., xxvii, 149, 1905. 




form ^(211) and jp (311). 



Ford — Neptmiite Crystals from. San Benito Co., Cat. 237 

observed this face was dull in luster and distinctly curved, and 
only approximate measurements could be obtained from it. 
Its symbol was determined, however, by its zonal relations, for 
it was found to lie at the intersection of the pyramid zone 
a (100) -s (111) -a (111) -i?(311) -a (100) and the zone m (110) 
— ^(112) — m (110). This form g is new to the mineral, for 
although Boggild in_Mineralogia Groenlandica, page 506, lists 
such a form as p (211) it is evidently a misprint, for in the 
original article by Wallenstrom* the form referred to is given 
as p (311). The various forms were identified by the follow- 
ing angles : 





Meas. 


Calc. 


a (100) /s jn(llO) 


49° 52' 


49° 53' 


a (100) /v c (001) 


64° 22^' 


64° 22' 


01(110) ^ s (111) 


38° 0' 


37° 58' 


02(100) /s c (001) 


73° 49' 


73° 49' 


02(110) ^ i (112) 


78° 26' 


78° 23V 


01(110) /s o (111) 


55° 14' 


54° 58' 


rn(llO) ^ g (211) 


40° 2' 


40° 28^' 


a (100) /v p (311) 


39° 38^ 


39° 27' 


igure 5 shows in stereographic projection 


all the crystal 




hiqioy 



mO'o) 



a ooo) 



forms of neptunite that have been listed. Those observed on 
the Greenland crystals are indicated by solid dots, while the 

* Loc. cit. 



238 Ford — Nept unite Crystals from San Benito Co., Col. 

forms found on the specimens from California are indicated by 
open circles. 

Optical. — For the optical study of the material a thin sec- 
tion was first cut parallel to the symmetry plane (010). On 
examination between crossed nicols it was found that one 
extinction direction was inclined to the c-axis of about 24° in 
the obtuse angle /3, while the second direction of extinction 
very nearly coincided with the «-axis making an angle of 1° 38' 
above it in front. The direction of greatest elasticity of the 
section corresponds to the extinction direction which nearly 
coincides with the a-axis. The section showed very strong 
pleochroism, the ray vibrating parallel to the direction a being 
colored yellow while the one vibrating parallel to c was a deep 
brownish red. These relations are shown in figure 6. Two 
other sections were cut, each being normal to one of the 



Fig. 6. 



Fig. 7. 



Fig. 8. 






ted 




red 



-0- 



red 



yello 



extinction directions of the section parallel to o (010). In the 
section normal to the ct-direction, or nearly normal to the a 
crystallographic axis, it was noted that the extinction direction 
which was parallel to the J-axis was the direction of greatest 
elasticity in the section or the direction of intermediate elas- 
ticity b of the mineral. This section showed in convergent 
polarized light an interference figure with a large axial angle, 
the optical axes lying in the symmetry plane of the crystal. 
This section showed no marked pleochroism, rays vibrating 
parallel to the two extinction directions being both colored a 
dark red. These relations are expressed in figure 7. 

The third section was cut normal to the direction c deter- 
mined in the section parallel to b (010) and in its orientation 
was nearly parallel to c(001). This section showed an inter- 
ference figure with the optical axes lying in the symmetry 



Ford — Neptunite Crystals from San Benito Co., Cat. 239 

plane of the crystal. As the angle between them was small 
it follows that the direction c, the normal to the section, is the 
acute bisectrix and the mineral is positive. Rays vibrating 
parallel to the o crystallographic axis were colored red, while 
those vibrating in the symmetry plane were yellow. These 
relations are shown in figure 8. 

The above results differ from those given in the original 
description in the position assigned to the axial plane. Flink 
describes the optical orientation of the Greenland neptunite as 
follows :* " The plane of the optical axes is perpendicular to 
the symmetry plane and the acute bisectrix forms an angle of 
18° with the vertical axis in the obtuse angle /3." A critical 
reading of his description would suggest that the discrepancy 
noted above was probably brought about by an accidental turn- 
ing of one of his sections so that the true orientation was 
reversed. To translate further from his description : " In the 
first section orientated parallel to the symmetry plane one extinc- 
tion direction forms with the vertical axis an angle of 18° in 
the obtuse angle (3. The rays vibrating in this direction show 
the greatest absorption and the color is deep red brown. 
The rays vibrating perpendicularly to this direction are less 
absorbed and the color is yellow-red. The same section shows 
also in convergent polarized light an axial figure with a large 
angle between the optical axes." With the exception of the 
last statement the description agrees with that given above for 
the California mineral. Flink describes the second section 
which he cut as follows : u In the second section, orientated 
perpendicular to the symmetry plane, and making an angle of 
72° with the orthopinacoid and 7J° with the base, the rays 
vibrating parallel to the symmetry axis are the least absorbed 
and the color is yellow-red. The rays vibrating perpendicu- 
larly to the symmetry axis and in the symmetry plane are more 
strongly absorbed and the color is dark red. This section 
shows in convergent polarized light an axial figure with a 
small angle between the optical axes." It will be noted that 
the elasticity direction which is common to the two sections, 
namely the one lying in the symmetry plane and making a 
small angle with the a-axis, is said in the first case to show the 
least absorption and to be colored yellow-red ; and in the 
second case the greatest absorption with a dark red color. 
This discrepancy could be explained and the optical orienta- 
tion of the Greenland mineral be brought into conformity with 
that from California, if Flink' s second section was considered 
to have been accidentally turned 90° from its true orientation. 
The statement that an axial figure with large optical angle was 
*Zeitschr. Ki\, xxiii, 350, 1894. 



24:0 Ford — Neptunite Crystals from San Benito Co., Col. 

obtained from the section parallel to (010) might be explained 
by the fact that often in biaxial crystals what appears to be an 
interference figure with a large axial angle can be obtained 
from sections that are cut parallel to the axial plane. 

Attempts were made to measure the indices of refraction of 
the mineral, but on account of the dark color of the material 
they failed. The average index of refraction was obtained in 
the following manner : Small fragments were observed under 
the microscope immersed in various mixtures of a monobrom- 
napthalene and methylene iodide until one was obtained which 
by use of the Becke method was found to be very closely of 
the same index of refraction as the neptunite. The index of 
refraction of the liquid was then determined by immersing in 
it fragments of other minerals with knowm indices of refrac- 
tion until the same result was obtained. In this way the 
average index of refraction of the neptunite was found to be 
close to 1*70. 

An approximate measurement of the axial angle was made 
by immersing the section cut perpendicular to the acute bisec- 
trix in a monobrom-napthalene (^=1*6583) and using a small 
axial angle goniometer placed on the stage of the microscope. 
The measured angle (2H) was approximately 50°, which w T ould 
give 2E = 90°. Assuming that /3=1*7, this would give the 
axial angle 2Y=48°40 / . Dispersion of the optical axes was 
noted, v >p. 

Chemical. — A qualitative examination of the California 
neptunite shows its substantial agreement with the analyses 
given of the Greenland material. It is hoped shortly to be 
able to make a quantitative analysis and to present its results 
in a subsequent paper. 

Mineralogical Laboratory of the Sheffield Scientific School of Yale Uni- 
versity, New Haven, Conn., Jan. 29, 1909. 



Gooch and Bosworth — Silver as the Chr ornate. 241 



Art. XYII. — The Gravimetric Determination of Silver as the 
Chromate ; by F. A. Gooch and Rowland S. Bosworth. 

[Contributions from the Kent Chemical Laboratory of Yale Univ. — cxcv.] 

It lias been shown in a recent paper from this laboratory* 
that the precipitation of silver chromate from the solution of 
a soluble chromate made faintly acid with acetic acid may be 
carried to completion by the addition of silver nitrate in con- 
siderable excess; and that the exact determination of the chrom- 
ium of a soluble chromate or dichromate may be effected by 
treating with silver nitrate the solution of either salt, adding 
ammonia to alkalinity and then acetic acid to faint acidity, 
transferring the precipitate and washing it in the filtering 
crucible with a dilute solution of silver nitrate until foreign 
material other than that reagent has been removed, finishing 
the washing with a small amount of water applied judiciously 
in portions, and weighing the dried or gently ignited residue 
of silver chromate. The success of this process turns upon 
keeping the chromium at the moment of precipitation essen- 
tially in the form of chromate rather than dichromate and in 
taking care that an excess of silver nitrate shall be present 
nearly to the end of the washing. The present paper deals 
with the conditions under which, in reversal of the process 
just described, silver may be precipitated completely as the 
chromate; 

In the first experiments to be described, a solution of silver 
dichromate was added gradually, either in slight excess or in 
considerable excess, to a solution of silver nitrate maintained 
at the boiling point ; the mixture was cooled and treated with 
ammonia until the precipitate first formed had been dissolved ; 
acetic acid was added to faint acidity ; and the precipitate 
was settled for a half -hour, filtered off on asbestos in a perfor- 
ated crucible, washed with a little water, dried and weighed. 
In every case the weight of silver chromate found was defic- 
ient and the testing of the filtrate with hydrochloric acid 
showed the presence of a soluble salt of silver. The errors 
amounted to several milligrams, varying with the conditions. 
Dilution of the original solution and prolonged washing with 
water increased the error. Admixture of alcohol was with- 
out appreciable eifect upon the solubility of the precipitate. 
Evaporation to dryness and transfer by a dilute solution of 
potassium dichromate, and a final washing with water used 
judiciously in small portions, tended to diminish the amount of 
soluble salt, but this treatment never reduced the error of loss 

* Gooch. and Weed : this Journal, xxvi, 85, 1908. 



242 Gooch and Bosworth — Silver as the Chromate. 

below two or three milligrams for volumes approximating a 
hundred cubic centimeters. Solution of the first precipitate 
in ammonia and reprecipitation by boiling to low volume to 
remove the excess of ammonia failed to overcome entirely the 
solubility of the silver salt, which appears to be due to a tend- 
ency on the part of the ammonium chromate to pass to the 
condition of dichromate with loss of ammonia from the boiling 
solution. 

In the next experiments, therefore, potassium chromate was 
used as the precipitant. In Table I, A are given the details 
of experiments in which an excess of potassium chromate was 
added to silver nitrate, the solution boiled, the precipitate 
transferred to the asbestos filter in a perforated crucible 
by means of a dilute solution of potassium chromate and 
washed with small portions of water, and the residue after 
drying with gentle heat weighed as silver chromate. In these 









Table I. 






Silver 
taken as 

AgN0 3 


K 2 Cr0 4 Ag 2 Cr0 4 
used weighed 


Silver 
found 


Error 
in terms 
of silver 


Volume 
of solu- 
tion 


Weight 

of 
silver 


Volume 
of solu- 
tion 


Weight 






cm 3 . 


grm. 


cm 3 . 


grm. grm. 
A 
ipitation by K 2 Cr0 4 . 


grm. 


grm. 






Prec 






15 


0-1652 


50 


0-3 0-2536 


0-1649 


— 0003 


10 


o-noi 


50 


0-3 0-1693 


0-1101 


o-oooo 


25 


0-1437 


50 


0*3 0-2200 


0-1436 


—o-oooi 


25 


0-1437 


50 


0-3 0-2210 


0-1437 


o-oooo 


Precipitat 


JD 

ion by K 2 CrOi, treatment with NHiOHan 


id evaporation to dryness. 


25 


0-1348 


50 


0-3 0.2077 


0-1351 


+ 0-0003 


30 


0-1618 


50 


0-3 0-2500 


0-1626 


+ 0-0008 


30 


0-1618 


50 


0-3 0-2520 


0-1639 


+ 0-0021 


Precipitat 


ion by K 2 CrOi, treatment with NH±OH and boiling to 

10-15 cmZ . 


a volume of 


25 


0-1576 


50 


03 0-2422 


0-1575 


—o-oooi 


25 


0-1576 


50 


0-3 0-2414 


0-1570 


— 0-0006 


50 


0-3152 


50 


0-3 0-4852 


0-3155 


+ 0-0003 


50 


0-3152 


50 


0-3 0-4843 


0-3149 


— 0*0003 


50 


0-3152 


50 


0-3 0*4847 


0-3152 


o-oooo 



D 

Precipitation by K 2 CrO iy treatment with NHiOH and boiling to a volume of 
10—15 cmZ in presence of 1 grm. of sodium nitrate. 

10 0-1101 50 0-3 0-1693 O'llOl 0*0000 
15 0-1652 50 0-3 0*2536 0-1649 — 0'0003 



Gooch and Bosworth — Silver as the Chromate. 



243 



experiments no silver was found in the nitrate, but the pre- 
cipitate was not so coarsely crystalline and easily washed as 
when deposited by removal of ammonia from an ammonia- 
cal solution. In the other experiments of Table I, therefore, 
the first precipitate was dissolved in ammonia and reprecipi- 
tated by boiling the solution : in those of B the evaporation 
was carried to dryness, and in those of C the concentration 
was stopped when a volume of 10-15 cm3 had been reached. 
In all these experiments precipitation was complete, the fil- 
trates giving no test with hydrochloric acid for a dissolved sil- 
ver salt. 

The high results obtained in the experiments of B are no 
doubt due to the inclusion of foreign matter in the silver chro- 
mate, which is left in caked condition by the process of complete 
evaporation. The better results in the experiments of C are 
apparently due to the fact that in them the evaporation was 
not pushed too far. The experiments of D show that the pres- 
ence of a gram of sodium nitrate has no appreciable effect 
upon the solubility of the silver chromate. It is plain that 
the precipitation by potassium chromate in neutral solution is 
practically complete and that accurate determinations may be 
made by filtering at once, or by dissolving the precipitate in 
ammonia, reprecipitating by boiling to a volume of 10-1 5 cm3 , 
and then filtering, drying, and weighing. 

In many cases it is desirable to determine silver present -in 
solutions containing free nitric acid. t The effect of free nitric 
acid upon the process was therefore next studied. A few 
qualitative experiments showed that the solvent action of nitric 
acid may be obviated by taking care to use the precipitant, 
potassium chromate, in such amounts that an excess of it shall 
be present after taking up the nitric acid to form potassium 
dichromate. The details of a few of these experiments are 
giving in the accompanying table. 









Table II. 












• 


K 2 CrO 


4 












To form 


Theo- 


s 




Agin 




To form 


K 2 O 2 7 


retically 




Ag 


AgN0 3 


HN0 3 


AgaCrO* 


with HNO3 


required 


Present 


in filtrate 


grin. 


grm. 


grm. 


grm. 


grm. 


grm. 




0-1376 


0-063 


0-1241 


0-1946 


0-3187 


0-3172 


Found 


0-1376 


0-063 


0-1241 


0-1946 


0-3187 


0-3172 


Found 


0-0550 


0-063 


0-0496 


0-1946 


0-2442 


0-3172 


None 


0-0550 


0-063 


0-0496 


0-1946 


0-2442 


0-3172 


None 


0-1376 


0-063 


0*1241 


0-1946 


0-3187 


0-3806 


None 


0-1376 


0-063 


0-1241 


0-1946 


0-3187 


0-4758 


None 


0*1376 


0-095 


0-1241 


0-2919 


0-4160 


0*4758 


None 


0-1376 


0-126 


0-1241 


0-3892 


0-5133 


0-6344 


None 


0-1376 


0-158 


0-1241 


0-4865 


0-6106 


0-7930 


None 



244 Gooch and Bosworth — Silver as the Chr ornate. 

In Table III are given the results of quantitative experi- 
ments in which precipitation was effected in presence of nitric 
acid, with care to insure the necessary excess of potassium 
chromate. In the experiments of A the precipitate was fil- 
tered off at once, without solution by ammonia and reprecipita- 
tion by boiling : in those of B the ammonia treatment was 
made to convert the less crystalline precipitate to better form 
for filtration and washing. 









Table III. 








Silver taken K 2 Cr0 4 










as AgN0 3 used 










Volume 


Volume 




HN0 3 






Error 


of solu- 


of solu- 




r * s 


Ag 2 Cr0 4 


Silver 


in terms 


tion 


Weight tion Weight 


Vol. Weight 


weighed 


found 


of silver 


cm. 


grm. cm. 


grm. 


cm 3 . grm. 

A 
KiCrOi in prese 


grm. 


grm. 


grm. 




Precipitat 


ion by 


nee of HN0 3 . 




25 


0-1355 50 


0-9 


10 0-182 


0-2091 


0-1360 


+ 0-0005 


25 


0-1355 50 


0-9 


10 0-182 


0-2081 


0-1353 


— 0-0002 


25 


0*1358 50 


0'9 


10 0-182 


02090 


0-1360 


+ 0-0005 


25 


0-1355 50 


0-9 


10 0-182 


0-2075 


0-1349 


— 0-0006 


25 


0-1355 50 


0-9 


10 0-182 


0-2090 


0-1360 


+ 0-0005 



Precipitation by K 2 CrOi in presence of HNO3, treatment with NHiOH, and boil- 
ing to a volume of 10-15 cm3 . 



25 


0-1348 


50 


0-6 


10 


0-063 


0-2076 


0-1350 


-fO-0002 


25 


0-1348 


50 


0-6 


10 


0-063 


0-2068 


0-1344 


— 0-0004 


25 


01348 


50 


0-6 


10 


0-063 


0-2072 


01347 


—o-oooi 


25 


0-1348 


50 


0-6 


10 


0-063 


0-2074 


0-1348 


0-0000 


25 


0-1348 


50 


0-6 


10 


0-063 


0-2070 


0-1346 


-0-0002 



So it appears that from solutions of silver nitrate, containing 
free nitric acid, potassium chromate precipitates silver chromate 
completely, provided enough potassium chromate is present to 
take up the nitric acid with formation of potassium dichromate 
as well as to form the silver salt. The precipitate, filtered at 
once or brought to better crystalline condition by treatment 
with ammonia and boiling of the solution to small volume, 
may be transferred to the asbestos filter by dilute potassium 
chromate and washed by small portions of water without 
appreciable loss. The weight of the residue of silver chromate, 
dried at gentle heat, may be taken as a very fair measure of 
the silver originally present. 




J. Trowbridge — Doppler Effect in Positive Rays. 245 



Aet. XVIII.— The Doppler Effect in Positive Bays ; by 
John Trowbridge. 

The discovery of canal rays by Goldstein, and that of the 
Doppler effect in these rays, marks an epoch in the study of 
the discharge of electricity through gases; for before these 
discoveries the multitude of confusing effects which arise in 
the space between the anode and the _, 1 

cathode made it difficult to observe 
any translation movement. The 
space, however, behind the cathode 
is comparatively free for the passage 
of the positive ions. 

We now recognize, in addition to 
the positive rays behind the cathode — 
the canal ray — retrograde positive rays 
which are directed to the anode, or 
rather away from the cathode in the 
direction of the anode.* This later 
discovery leads one to expect that the 
Doppler effect should be found also 
between the anode and the cathode. The result of my study 
shows that the effect does exist in this region and indicates 
a movement away from the cathode and toward the anode. 

The form of tube I have employed is represented in fig. 1. 
The slit of the Rowland grating was at X for the retrograde 
rays and at Y for the canal rays; A being the anode and C 
the cathode. The Rowland grating gave, in the order of spec- 
trum I employed, six Angstrom units to nine-tenths of a milli- 
meter. The effect was observed with respect to the hydrogen 
line 4861*5 and the change in refrangibility was measured by 
comparison with the solar spectrum, which was photographed 
immediately beneath the gaseous line without changing the jaws 
of the slit. The amount of the change in refrangibility was 
sensibly the same as in the canal rays. 

The difference of potential between the anode and the cathode 
varied between five thousand and ten thousand volts ; and the 
current from ten to five milliamperes furnished by a storage 
battery of ten thousand cells. A current of running water 
provided a large and steady resistance. 

The appearance of the discharge at the cathode has often 
been described. The cathode appears to be the base of two 

* Wehnelt, Wied. Ann., 1899, p. 421 ; Eunge and Baschen, Wjed. Ann., 
lxi, 1897, p. 644; Paschen, Wied. Ann., xxiii, p. 247; Villard, Comptes 
Kendus, cxliii, p. 673, 1906; Goldstein, Phil. Mag.. March, 1906; Jacob 
Kunz, Phil. Mag., July, 1908, p. 161 ; J. J. Thomson, Oct., 1908, p. 657. 

Am. Jour. Sci. — Fourth Series, Vol. XXVII, No. 159. 
17 



246 J. Trowbridge — Dojppler Effect in Positive Bays. 

rose-colored cones of light, the apex of one directed to the 
anode and the apex of the other toward the canal region. The 
body of the luminous cone in the space between the anode and 
the cathode is, so to speak, a solid, while that in the canal 
region, or back of the cathode, is made up of a collection of 
tubes which in a short region come together at the apex of the 
cone, and in a more extended region spread out in a diffused 
manner. When the cathode is unperforated the rosy glow, 
which in the case of hydrogen characterizes the canal rays, 
emanates from the central portion of the aluminium cathode ; 
it is no longer conical in form, or rather resembles a frustrum 
of a cone, the base directed to the anode. It does not extend 
as far toward the anode as the conical discharge from the per- 
forated cathode and is not so bright. With the unperforated 
cathode and the slit placed at X, no Doppler effect was seen. 
When, however, the cathode was perforated the effect was very 

Fig. 2. 




. j 



evident and indicated a movement toward X which was equal in 
amount to that observed in the canal region toward Y. More- 
over, the photographs showed, when the light was observed at 
X, a line on each side of the ordinary stationary hydrogen line ; — 
my observations were confined to 4861*5. There was evidently 
a movement toward the anode, and a movement away from it 
at the cathode. 

When the observations were conducted at Y, the same 
phenomenon was observed : a stationary hydrogen line and a 
diffuse line separated from the stationary line by a blank space 
on each side of the stationary line — indicating a movement 
toward the cathode and away from it. In fig. 2, a and b are 
photographs, o is a drawing which represents effects too feeble 
to be strongly reproduced from the photographs — effects, how- 
ever, which are very evident on the negatives a and b. The 



J. Trowbridge — Doppler Effect in Positive Rays. 247 

slit was a broad one in order to show differences of illumi- 
nation. The light was strongest at the orifices. 

When the observations were conducted by placing the slit 
of the spectroscope so that the light at the perforations did 
not enter the slit — in other words, placing it obliquely to the 
band of light — the companion of the stationary line which 
indicated a movement away in each case, from the slit, was 
not discernible. The effect took place at the orifices. The 
positive particles jostling' through these orifices and mutu- 
ally repellent transmit movenfents, like those resulting from 
elastic particles in impact, in opposite directions and those 
driven in the direction of the anode meet others coming 
toward the cathode. There resnlts a maximum of radiation 
of greater refrangibility which is separated from the refrangi- 
bility of the stationary hydrogen line by a less luminous space. 

When glass tubes are inserted in the orifices through which 
the canal rays pass and the back of the cathode is protected 
by glass connected to these tubes and to the wall of the dis- 
charge tube, the canal rays are still obtained. This proves that 
these rays are produced immediately in front of the cathode 
that is on the side toward the anode, or in the orifices. I 
incline to the belief, as 1 have stated, that the jostling in the 
narrow orifices accounts for the change in refrangibility. 

In all discharge tubes striae are seen opposite to the edge of 
the cathode on the glass. These striae can be localized at a 
definite point by bringing the edge of the disc of the cathode 
near the wall of the tube ; and we then have a source of positive 
rays which is analogous to that formed by pushing a glass 
tube surrounding the anode into the Crookes space.* 

The heating and oxidizing effect from the positive particles 
of these striae is very marked, especially when the canal region 
is small, and is much greater than any effect produced by the 
canal rays in a possible rebounding from the end of the canal 
region. The photographs of cathode discs exhibit this effect 
of the positive rays coming from the striae. In fig. 3, No. I is 
a photogi^h of the back of a cathode made of aluminium 
formed from as pure clay as could be obtained. The disc was 
•T5 mm thick and the hard surface formed by the iron rolls was 
left upon it. The heat from the positive rays coming from 
the striae caused the occluded gases of the aluminium to form 
blisters on its surface. No. II represents the back of a cathode 
formed from the same quality of aluminium as in case I, 
except that the plate of aluminium was treated with nitric acid to 
remove any trace of iron coming from the rolls. The surface 
thus lost its polish and hardness. A very black deposit formed 

*E. Wiedemann, Wied. Ann., lxiii, p. 242, 1897. 



248 J. Trowbridge — Doppler Effect in Positive Bays. 

under the effect of the positive rays of the striae, which was 
probably carbon from the aluminium. In III the cathode was 
formed from ordinary commercial aluminium l*5 mm thick. It 
represents the blackening which results from long running of 



Fig. 3. 




II 



ill 



IV 



Fig. 4. 




the] discharge. At increased exhaustions and during this 
long use of the tube the striae shift their position, and the back 
of the cathode shows a general discoloration. No. IV repre- 



J. Trowbridge — Doppler Effect in Positive Rays. 249 

sents the front of the cathode of fig. 3. The center is bright, 
while the edges are discolored. Goldstein has noticed that 
positive rays remove deposits. 

Fig. 4 is a photograph of the canal rajs which shows also 
the luminosity on the back of the cathode produced mainly by 
the strige. 

The order of spectrum produced by the Rowland grating; 
gave an interval of '9 millimeter to six Angstrom units. The 
approximate interval between the Doppler effect on both sides 
of the stationary hydrogen line 4861'5 was three Angstrom 
units ; the difference of potential between the anode and the 
cathode varied from 6000 volts to 10,000. The current ran 
from 10 milliamperes to 5. A storage battery of 10,000 cells 
was emploved. The internal diameter of the discharge tubes 
was 3 cm . the distance between X and C, fig. 1, was 6 cm . The 
distance between C and Y varied from 4 cm to 10. 

Jefferson Physical Laboratory, 

Harvard University, Cambridge, Mass. 



250 G. R. Wi eland — Armored Saurian from the Niobrara. 

Art. XIX. — A New Armored Saurian from the Niobrara ; 
by G-. R. Wieland. 

The greatest American storehouse of fossil marine verte- 
brates is doubtless in the Niobrara chalk of western Kansas. 
But despite the fact that many of the diverse forms there rep- 
resented must have lived near and frequented the shores of 
the Niobrara sea, very little evidence of even presumably true 
land forms has thus far been obtained. 

The best known form to be regarded as land, or at least 
lacustrine or fluviatile, is the Hadrosaurus agilis of Marsh 
from the Smoky Hill river. Though the type of this dino- 
saur includes considerable portions of the skeleton, only a single 
individual has ever been recovered. In fact, in the University 
Geological Survey of Kansas, vol. iv, Professor Williston says, 
in speaking of the Dinosauria : " But a single specimen (that 
is Hadrosaurus agilis) has ever been found in the state, so far 
as I am aware, though the animals must have lived here about 
the shores of the Cretaceous seas in great abundance." 

And although collecting in the Niobrara has been especially 
active during the past ten years, no further examples even 
doubtfully referable to Dinosauria came within my knowledge 
until about two years ago. Then [ noted amongst turtle 
material sent to the Yale Museum from the Hackberry Creek 
region by Mr. Charles H. Sternberg, and referred to me for 
study, the two paired and presumably caudal, or else cervical, 
dermal elements shown in figure 7, 7a. 

On the basis of such slender evidence it was of course not 
possible to say whether a crocodilian was indicated, or even 
some remote progenitor of such a turtle as Miolania. Now, 
however, we are enabled to present some clearer evidence for 
the presence of a second Dinosaurian genus in the Niobrara. 

This last season Mr. Sternberg secured, five miles south of 
Castle Rock and three miles south of Hackberry Creek, six 
dermal scutes of a form quite certainly dinosaurian. These 
plates later came into my possession and have been donated to 
the Yale Museum. The name Hierosaurus Stembergii is 
assigned them in honor of their Collector. It is thought that 
other fragmentary specimens have been observed, so that it is 
probable that further material will yet be obtained. 

In the excellent figures l-7a of the present fossil scutes, 
drawn by Mr. R. Weber, the principal characters may at once 
be discerned. 

The scutes are all shown one-third the natural size in the 
figures l-7a. The bones shown in figures 1-3 are odd, that 
of No. 3 being merely a tubercle with a fine right-angled stria- 
tum on its lower surface. Those shown in Nos. 4 and 5 are a 
pair, but other elements must have intervened ; while the two 
fused elements shown in No. 6 form an isolated asymmetrical 



G. R. Wieland — Armored Saurian from the Niobrara. 251 

Figs. l-7a. 




252 G. B. Wieland — Armored Saurian from the Niobrara. 

plate. In all the foregoing except the tubercle the thickness 
is much as seen in the transverse section No. 2Z>. The more 
or less median ridge is sharp and runs the entire length, being 
of much the same height throughout its course, and terminat- 
ing as a sharp backwardly projecting slightly upturned spur. 
The height of the ridge is from one to one and a half centi- 
meters. Eos. 1-6 are all plate-like and of much the same 
thickness as shown in the middle transverse section figure 2h. 

The two elements shown in figure 7, 7a are probably a 
terminal pair that was seated on the proximal caudal or cervi- 
cal region, as indicated by the broad flat front edge which 
formed a contact and the free thinner posterior edge which 
appears to end [if not begin] the series abruptly. But the ani- 
mal to wmich these odd bones belonged may not even have 
been of the same species as that to which the scales of figures 
1-6 belong and form the type. 

It is pretty clear that the form before us is allied to the 
Stegosau ridse and is possibly included in the Ankylosauridse of 
Brown* represented by Dinosaurs with large shields and a 
quite rigid turtle- like back from the Judith River beds near 
Gilbert Creek, 120 miles north of Miles City, Montana. 

The closest relationship within the family Ankylosauridae so 
far as the dermal armature affords comparison is afforded by 
Polacanthus as restored by JNopcsa and now on exhibition in 
the British Museum at South Kensington. As there restored 
there are first free plates and then a more or less perfectly 
developed carapace only extended as such over the lumbar and 
hip region. 

From the fact that we see the large scales shown in figure 6 
so completely fused a lumbar-hip carapace may be supposed 
present in the Niobrara form. But in the latter there is plainly 
indicated by heavy sulci the presence of a system of horn- 
shields at least as large as the keeled plates. These characters 
evidently form a sufficient generic distinction. 

It thus seems probable that the Dinosaurs actually paralleled 
the turtles in the development of keels of dermogene bones 
enclosed by horny shields and coming near to the formation of 
a true carapace with a clearly aligned bone and hornshield 
system primarily comparable to that of Dermochelys as it now 
exists. It is not to be forgotten, however, that the unusual 
structure of the Archelon carapace described in previous pages 
makes it very likely that ere long a turtle may be found with 
a neural, pleural and marginal series greatly reduced and 
mainly replaced by rows of large shields not greatly unlike 
those now described. 

Yale University Museum, New Haven, Conn. 

* The Ankylosauridae, a new family of Armored Dinosaurs from the Upper 
Cretaceous ; by Barnum Brown. Bull. Amer. Mus. Nat. Hist., vol. xxiv, 
Art. XII, pp. 187-201, New York, Feb. 13. 1908. [Though I fail to see why 
the Nodosauridae of Marsh are ignored in this paper.] 



Wolcott Gills. 253 



WOLCOTT GIBBS. 

Wolcott Gibbs, for many years an associate editor of this 
Journal, and during the last part of his scientific career the 
most commanding figure in American chemistry, was born in 
New York, February 21, 1822. 'His father, Colonel George 
Gibbs, was one of the earliest American mineralogists, and is 
commemorated in the mineral Gibbsite. He was a friend of 
the elder Silliman, and his fine collection, deposited in New 
Haven in 1812 and purchased in 1825, became the foundation 
of the mineral cabinet of Yale College. His mother, Laura 
Wolcott Gibbs, was the daughter of Oliver Wolcott, Secretary 
of the Treasury during part of the administrations of Wash- 
ington and John Adams, and granddaughter of the signer of 
the Declaration of Independence of the same name. The 
child, who was the second son, was named Oliver Wolcott 
Gibbs, but, as he disliked the name of Oliver, he dropped it 
in early life, and is known to the scientific world as Wolcott 
Gibbs. 

The taste for science inherited from his father was not slow 
in appearing, for, as he tells us, even in his early childhood, 
which was passed mostly at his father's large estate called 
Sunswick on Long Island a few miles from New York, " he 
was often occupied with making volcanoes with such materials 
as he could obtain, and in searching the stone walls on the 
estate for minerals and the gardens and fields for flowers." 

At the age of seven he went to live with William Ellery 
Channing, the great Unitarian divine, who had married his 
aunt, but he was under the special care of another aunt, Miss 
Sarah Gibbs. The winters were passed in a fine house on Mt. 
Vernon street, Boston, and the summers at Oakland, a beauti- 
ful estate about five miles from Newport, R. I. The fame of 
Dr. Channing brought many foreign visitors, especially in sum- 
mer, and this stimulating mental atmosphere, to which the 
boy was exposed for Hve years, had a marked effect on his 
intellectual development. 

In 1837 he entered Columbia College, and his first original 
work dates from his junior year there. It consisted of a new 



254 Wolcott Gibbs. 

form of galvanic battery, in which carbon was used for the 
first time as the inactive plate. Upon receiving his degree of 
A.B. in 1841 he began his chemical education by taking a 
place as assistant with Dr. Robert Hare, professor of chemis- 
try in xhe Medical School of the University of Pennsylvania. 
His long life, therefore, linked one of onr earliest chemists 
with those of the present generation. After some months in 
this laboratory he entered the College of Physicians and Sur- 
geons of New York with the intention of qualifying himself 
to hold the chair of chemistry in a medical school. It is cer- 
tainly remarkable that at this early day he should have been 
able to work out so well-conceived a plan for chemical study ; 
which he continued, after taking the degree of M.D. in 1845 
and incidentally the A.M. in 1844, by going to Europe, where 
he entered the laboratory of Pammelsberg at Berlin. Some 
months here were followed by a year under Heinrich Pose, 
also in Berlin, a semester in Giessen with Liebig, and courses 
of lectures in Paris from Laurent Dumas and Pegnault. Of 
this brilliant constellation of teachers, Heinrich Pose had by 
far the most influence on him, as is shown by his lifelong 
devotion to inorganic and analytical chemistry in spite of the 
fascinations of organic chemistry under Liebig and physical 
chemistry under Pegnault. He always spoke of Pose with 
the greatest admiration and affection, and evidently regarded 
him as his chemical father. 

In 1848 he returned to New York, and, after giving a short 
course of lectures at Delaware College, Newark, was elected in 
1849 professor of chemistry and physics in the newly created 
Free Academy now called the College of the City of New 
York. For the next eight years there is little to record except 
his marriage to Josephine Mauran in 1853, and the fact that 
in 1851 he began a series of reports on chemical and physical 
progress for this Journal as associate editor, continued till 18Y3, 
which form 500 pages of succinct but clear and comprehensive 
abstracts of the most important papers of this period. He also 
carried on a similar work for American chemistry as corre- 
spondent of the German Chemical Society from 1869-1877. 
Until 1857 his papers, few in number and of no great import- 
ance, only show he was. finding his feet, but in that year 



Wolcott Gills. 255 

appeared the account of his great investigation of the ammonia- 
cobalt compounds with F. A. Genth, which contained so full 
and thorough a study of the principal series of these puzzling 
bodies that very little in the way of experiment was left for 
future work, the analytical and chemical work being supple- 
mented in many cases by crystallographic determinations by 
J. D. Dana. So exhaustive a research was a new thing in 
American chemistry, and at once established his reputation on 
a firm basis. 

In 1861 appeared the first of three papers on new methods 
of separating the platinum metals, an important research on a 
most difficult subject, which was a worthy companion of his 
great work on the ammonia-cobalt compounds. These investi- 
gations led to his election in 1863 to the Kumford Professor- 
ship of the Application of Science to the Useful Arts in 
Harvard University, left vacant by the retirement of Professor 
E. 2sT. Horsford, and accordingly, after he had established 
himself in Cambridge, he took charge of the chemical labora- 
tory of the Lawrence Scientific School. The number of his 
students was small, but more were not to be expected or 
desired, as the object of the course was to educate professional 
chemists, and the supply was somewhat greater than the 
demand. His own work during this period is described in a 
number of short papers principally on chemical analysis ; the 
most conspicuous of which, introducing the electrical deposi- 
tion of the metals as a means of their quantitative determina- 
tion, laid the foundation of what has since become a new 
department of the science — electrical analysis. Another of 
these papers on the sand or glass filter is interesting as a fore- 
runner of the Gooch crucible, and his experimental method for 
correcting the volumes of gases may also be mentioned. 

In 1871 a reorganization of the Chemical Department con- 
solidated the laboratory of the Lawrence Scientific School with 
that of Harvard College, and relegated Dr. Gibbs to the 
Department of Physics, where he taught a small advanced 
class in Light and Heat. Some papers on optical subjects 
inspired by the discovery of the spectroscope probably led to 
this assignment, which was justified on the score of economy, 
but its wisdom may be doubted, as it deprived the chemical 



256 Wolcott Gibba. 

students of the university of the teaching of the best chemist 
in the country, and diminished the volume of his original 
work, since up to a certain point the amount of chemical pro- 
duction is directly proportional to the number of hands at the 
disposal of the master. Yet a study of his papers shows that, 
when his time was occupied by the administration of a labora- 
tory and more elementary teaching, he did not produce those 
extended researches on which his fame principally rests, as 
these date from the earlier and later periods, when his whole 
energy was concentrated on work of his own with, in the later 
period, one skilled private assistant. 

His two earlier investigations of this sort had to do with 
subjects so abstruse and difficult that most chemists would have 
shuddered at the idea of attacking them, but, as he once said, 
he was a pioneer, and seemed to enjoy nothing more than 
breaking a way through these tangled jungles on the frontier 
of the science. Accordingly he next took up a field of work — 
the complex acids of tungsten and molybdenum — even more 
terrible, for here it takes courage merely to read his papers, 
and follow his footsteps through the bewildering maze of 
series after series of compounds. What then must it have 
been to find the necessary clue to this labyrinth, and to estab- 
lish the nature of these numerous compounds ? Especially 
since in doing this it was necessary for him to work out some 
of the most difficult problems of analytical chemistry, as the 
separation of many of the elements involved had never been 
attempted before. In this great investigation over fifty new 
series of compounds were discovered by him, and the old series 
fully investigated and put on a solid foundation. The first 
paper appeared in 1877, and the last in 1896, this being the 
last paper published by him. 

The invention of the ring burner, his most important con- 
tribution to the apparatus of analytical chemistry, dates from 
1873, just after his transfer to the Department of Physics ; and 
his porous diaphragms for heating precipitates in gases were 
described in the same year. In 1877 appeared his excellent 
method for preparing nitrogen. 

In 1887 he became professor emeritus, and retired to New- 
port, where he continued his work on the complex acids in a 



Wolcott Gills. 257 

laboratory which he built for the purpose ; and also invaded 
the only remaining department of inorganic chemistry, which 
could be ranked for complexity and difficulty with the three 
already occupied by him, — this was the cerium group, but 
advancing years prevented him from making more than a pre- 
liminary exploration in this Held. 

In these last years at Newport he also took up an extended 
study of the effect of isomeric organic compounds on animals, 
at first with Dr. H. A. Hare and later with Dr. E. T. Reichert. 
In addition to this work in physiological chemistry he had in 
his earlier papers made occasional excursions into theoretical 
chemistry, organic chemistry, mineralogy, and physics, an 
astonishingly broad field to be covered by one man. His fame, 
however, rests on his work in inorganic and analytical chem- 
istry, and it seems to me that his two qualities, which make 
this preeminent, are thoroughness and accuracy. Next to 
these I should place his wonderful suggestiveness. His ideas 
flowed rather in a torrent than a stream, and occasionally bore 
him away from some good subject, after he had little more 
than broken the ground in it. Even old age did not check 
this current, his paper on the cerium metals published at 72 
containing five new and ingenious suggestions for their sepa- 
ration. The criticism, which might be made on his work, is 
that he was content to prepare the experimental foundations, 
but left it to others to build the theoretical structures upon 
them. The reason for this, I think, was that before all else 
he was an experimentalist. His table was always covered with 
a multitude of test tubes each with a label in its mouth setting 
forth the substances reacting within, and it was beautiful to 
watch the sharpness of insight with which he seized on the 
favorable indications in these numerous experiments, and the 
rare skill and robust energy with which he followed any line of 
work that promised results of value. 

As a teacher he had an especial faculty for imbuing his stu- 
dents with the enthusiasm and spirit of original work ; and 
they felt the greatest admiration and affection for him. The 
only instruction I received from him was a single voluntary 
lecture in my senior year. In this his ideas came hurrying out 
with an impetuous speed, as if there were too many to be 



258 Wolcott Gills. 

forced into the narrow limits of an hour. The effect was 
wonderfully inspiring, when, as in this case, one did not have 
to take notes. As already stated, his teaching was confined to 
a few students, and it is a matter of regret that more did not 
have the opportunity of profiting by contact with this great 
mind. 

Of the various societies, to which he belonged, he was most 
warmly attached to the National Academy of Sciences. He 
was one of its founders in 1870, and at his death the last sur- 
vivor of the original members. He served it as foreign secre- 
tary, vice president, and from 1895 to 1901 president. He 
was also a member of the American Academy of Arts and 
Sciences and of the American Association for the Advance- 
ment of Science, of which he was a vice president in 1866, 
president in 1897. In addition to these his achievements 
brought him honorary membership in the Philosophical Soci- 
ety of Philadelphia, and the American, English, and German 
Chemical Societies, and corresponding membership in the 
British Association for the Advancement of Science, and the 
Royal Prussian Academy. The most striking of these appre- 
ciations of his merit was his election as honorary member of 
the German Chemical Society, as he was the only American on 
this list. The degree of LL.D. was conferred on him by 
Columbia University, Harvard University, the University of 
Pennsylvania in alsentia as a special honor, and the Colum- 
bian University of Washington. 

During the Civil War he proved himself a public spirited, 
patriotic citizen, devoting a large part of his time to service on 
the executive committee of the Sanitary Commission. The 
frequent meetings of this body suggested to him the idea of 
a a club which should be devoted to the social organization of 
sentiments of loyalty to the Union." A meeting to consider 
this plan was called at his house on January 30, 1863, and led 
to the formation of the Union League Club of New York. In 
later life, although he took no prominent part in public affairs, 
he always held his knowledge at the service of the govern- 
ment, as was shown by an extensive report on the instruments 
for physical research, prepared when he was a commissioner 
to the exposition at Vienna in 1873, another on the tariff as 
applied to seeds, and others on various chemical subjects. 



Wolcott Gills. 259 

His most engrossing pursuit, next to chemistry, was garden- 
ing, a taste inherited from his father, and he threw himself 
into this with the enthusiasm, skill, thoroughness, and success, 
which characterized his work in chemistry. It served as an 
excellent relief from his arduous labors, and, when old age 
brought these to an end, as a delightful occupation, until a 
short time before his death, which took place at Newport, 
December 9, 1908. 

I wish I could bring before the reader the pictures which I 
like best to remember, when I think of him. The tall, hand- 
some man brimming over with warm, cordial welcome as he 
hastened to meet you on your arrival at his house, and later, on 
the piazza overlooking the garden and the sea, the long inspir- 
ing talks, which kept you in a high clear country far above 
anything mean or questionable, and sent you back to your 
work with renewed energy and enthusiasm ; his vivid enjoy- 
ment in the pursuit of his experiments in the laboratory ; or 
best of all, to see him wandering from bed to bed in his sunny 
garden, rejoicing in each rare and beautiful plant in his rich 
collection. 

C. L. Jackson. 



260 Scientific Intelligence. 



SCIENTIFIC INTELLIGENCE 



I. Chemistry and Physios. 

1. New Method of Forming Liquid Alloys of Sodium and 
Potassium. — It has been found by Jaubert that when metallic 
sodium acts upon potassium hydroxide, or when metallic potas- 
sium acts upon sodium hydroxide, there are formed in both 
cases, at a slightly elevated temperature, liquid alloys of the 
two alkali metals containing up to nearly 80 per cent of 
potassium. The very remarkable circumstance that both of 
these reactions take place and both give practically the same 
product appears to be due to the fact that these alkali metals 
combine to form the compounds NaK and NaK 2 with the libera- 
tion of a considerable amount of heat. The following equations 
represent the reactions : 



K, + NaOH = KOH + NaK,, 
Na 3 + 2KOH = 2NaOH + Nal 
Na„ + KOH = NaOH + NaK. 



These reactions give almost quantitative yields, and they are 
brought about by heating the reacting substances to from 225° 
to 350° in the different cases. The operation may be carried 
out either under paraffine or in an exhausted vessel. The 
reaction corresponding to the second of the equations given 
above is one by means of which an alloy rich in metallic potas- 
sium is now being made cheaply on a commercial scale in 
France, and the product promises to have important industrial 
applications, since, as is well known, potassium in many cases 
gives reactions entirely different from those of sodium, and it often 
reacts where sodium is without action. In connection with 
these results the author points out the falsity of the statement, 
given in most works on chemistry, that the alkali metals react 
with their hydroxides to form their oxides ; for example, 

Na + NaOH = Na 2 + H, 

and that only the opposite reaction, 

Na 2 + H =: NaOH + Na, 

is possible, as has been shown by Becketoff. — Bulletin, IV, iii, 
1126. h. l. w. 

2. The Determination of Cerium and other Rare Earths in 
Rocks. — It has been found by Dietrich that when oxalic acid 
or ammonium oxalate is added to a cerium solution in the pres- 
ence of ferrous sulphate, cerium oxalate is precipitated, but this 
precipitate contains a considerable amount of iron when a large 
quantity of this metal is present. When a neutral solution of a 
ferric salt is present no precipitate of cerium oxalate is produced 



Chemistry and Physics. 261 

until the iron has been converted into the green complex oxa- 
late, and then when an excess of oxalic acid or ammonium 
oxalate has been added the cerium oxalate is gradually precipi- 
tated in a pure white condition. Even when cerium oxalate has 
actually been precipitated it dissolves upon the addition of ferric 
chloride. The author advises, therefore, the use of a very much 
larger amount of ammonium oxalate than is customary in analyt- 
ical work when cerium is to be separated from iron. He prefers 
to use ammonium oxalate rather than oxalic acid for this pur- 
pose, as the latter makes the liquid too acid. When ammonium 
oxalate is employed in this way, however, all the calcium present 
is precipitated with the rare earth oxalates, but the latter may 
be separated by igniting the precipitate, dissolving the residue in 
acid and precipitating twice with ammonia. — Berichte, xli, 4373. 

h. l. w. 

3. Solubility of Metallic Gold in Hydrochloric Acid in the 
presence of Organic Substances. — But lew solvents for metallic 
gold have been known up to the present time. These are chlor- 
ine, bromine, iodine, cyanides, and selenic acid. Aweekiew 
has now observed that iinely divided metallic gold is soluble to 
some extent in hydrochloric acid in the presence of many organic 
substances. Among the organic compounds showing this pro- 
perty to a notable extent are the following, which are arranged 
according to diminishing effectiveness in this respect : methyl 
alcohol, amyl alcohol, chloroform, ethyl alcohol, chloral hydrate, 
phenol, cane sugar, glycerine, trioxymethylene, formaldehyde. 
The action takes place very slowly at ordinary temperature, but 
it is more rapid upon boiling, although the duration of the boil- 
ing appears to have -no essential influence. The gold must be 
very finely divided to show such solubility, and large amounts 
of solvent dissolve small amounts of gold. For instance, 1000 cc 
each of methyl alcohol and hydrochloric acid when boiled for 
5 hours with "1154 g. of gold precipitated with ferrous sulphate 
dissolved -0128 g. of it. The results indicate the existence of a 
new class of gold compounds, which it is the author's intention 
to investigate. — Zeitschr. anorgan. Chem., lxi, 13. h. l. w. 

4. The Composition of Matter. — The following contribution 
to theory has been put forth by E. Mulder : The conception 
that the atoms, which form the molecules, are compound has been 
advanced many times. But as yet he has not observed the 
hypothesis that these atoms of the second order are composed 
of atoms of the third order, and so on to infinity, a view which 
he advanced long ago. The ether, a particular and hypothetical 
form of the same matter (differing only in condensation), has, 
according to this idea, an analogous constitution. In the case 
of matter there are presented then, at the end of ends, infinitely 
small atoms, a conception not easily grasped. 

A theory such as this, which explains no known phenomena, 
appears to be unprofitable. Many others besides Mulder have 
doubtless indulged in just such a speculation, and it is easy to 

Am. Jour. Sci.— Fourth Series. Vol. XXVII, No. 159.— March, 1909. 

18 



262 Scientific Intelligence. 

carry the idea upward as well as downward ; to assume that the 
planets and fixed stars are atoms of the next higher order, that 
there are many orders on beyond, so that at the end of ends there 
would be infinitely large atoms, a conception which is likewise 
not easy to grasp. — Recueil, xxvii, 418. h. l. w. 

5. Introduction to the Rarer Elements, by Philip E. Brown- 
ing, 8vo, pp. 200. New York, 1908 (John Wiley & Sons). — 
This is the second edition, thoroughly revised and considerably 
enlarged, of a book which first appeared five years ago and is well 
known to the chemical public. Among the additions found in 
the new edition is a chapter by Dr. Boltwood, giving an excel- 
lent and concise account of the Radio-Elements. h. l. w. 

6. Feste Lo sung en und Isomorphismus, by Gitjseppi Bruni. 
12mo, pp. 127. Leipzig, 1908 (Akademische Verlagsgesell- 
schaft). — The author of this book has taken an active part in the 
development of modern views regarding solid solutions. In this 
volume, he gives an exceedingly clear account of the work that 
has been done and the results that have been obtained. The first 
part consists essentially of a lecture delivered at Breslau and this 
is supplemented with an appendix containing additional notes 
and references. h. w. f. 

7. The Production of Helium from Uranium. — F. Soddy 
refers to his previous work on the production of helium from 
thorium and extends his work by a study of the production of 
this gas from uranium. He concludes that this gas is produced 
from uranium with an approximate velocity of 2-10 -12 (year) -1 . 
That is, out of 1,000,000 kg. of uranium 2 mg. of helium are 
produced per year. Preliminary results with sylvine give a 
production velocity of 2*5 X10 -12 (year) -1 . — Physik. Zeitschrift, 
Jan. 15, 1909, p. 41. j. t. 

8. The Charge and Nature of the a- Particle. — E. Ruth- 
erford and H. Geiger review the work upon this subject, 
comparing the charge on the a-particle with that on the hydrogen 
atom — finding it between 2e and 3e. They compare various 
methods of determining this charge and discuss the accuracy 
of different methods of determination. It is maintained that 
the quantity of helium formed from a gram of radium is 
5-0xl0 -9ccm per second. The life of radium is 1760 years. — 
Physik. Zeitschrift, Jan. 15, 1909, pp. 42-46. J. t. 

9. Amount of Water in a Cloud formed by Expansion of 
Moist Air. — In the celebrated experiments of C. T. R.Wilson and 
Professor J. J. Thomson — on the determination of the number 
of ions by means of condensation of vapor — it is assumed that 
the air is cooled to the full extent by the adiabatic expansion 
before the drops begin to form. Professor W. B. Morton has 
investigated the conditions which would result if the expansion 
and condensation kept pace with each other, the process being a 
reversible one, and find that Thomson's results and his own 
differ only by six per cent. The assumption of complete adia- 
batic cooling of the air does not introduce an error comparable 



Geology and Natural History. 263 

with other inaccuracies in the experiment. — Phil. Mag., Jan., 
1909, pp. 190-192. J. t. 

10. Permanent Magnetism of Copper. — J. G. Gray and A. D. 
Ross have studied the behavior of pure copper in strong mag- 
netic fields under various conditions of temperature varying from 
bright red heat to that of liquid air. The traces of remanent 
magnetism are recorded. Ordinary and electrotytic copper were 
investigated together with the traces of iron. — Physikal. Zeit- 
schrift, Jan. 15, 1909, pp. 59-61. J. t. 

11. United States Magnetic Tables and Magnetic Charts, for 
1905 ; by L. A. Bauer. Pp. 154, 1 figure, 7 charts. Washing- 
ton, 1908 (Department of Commerce and Labor, Coast and Geo- 
detic Survey). — The latest results obtained for the magnetic 
elements of the United States and the immediately adjacent 
countries are presented together in this volume, which appears 
independently instead of being attached as an appendix to the 
reports of the Coast and Geodetic Survey. The author states 
that he resigned his position in connection with the survey on 
September 1, 1906, but that the matter was essentially complete 
at that date except as regards the determination of the correc- 
tions for secular variation. In order that these may be as accu- 
rate as possible, the charts now issued are made out for January 
1, 1905, instead of a subsequent date, as for 1910.* Recent 
experience has shown that unexpected changes have taken place 
in the variation and that they have not progressed in accordance 
with the predictions from the empirical formulae established 
earlier, hence the necessity of great care to insure accuracy. The 
number of stations in the country for which the magnetic ele- 
ments are tabulated. is 3311 (2869 C. and G. Survey Stations),. or, 
in other words, there is one magnetic station for about 973 square 
miles. Furthermore, magnetic work at sea was begun on Coast 
Survey vessels in 1903, so that the results obtained by them from 
1903 to 1907, as well as by the Carnegie Institution along the 
Pacific coast, have been available for this work. In addition to 
the various tables, the report includes seven charts giving the 
declination, inclination, horizontal intensity, vertical intensity, 
total intensity, also the magnetic meridians, and, finally, the secu- 
lar motion curves and horizontal intensity secular variation curves. 



II. Geology and Natural History. 

1. Die Fossilen Insekten and die Phylogenie der Pezenten 
Formen ; von Anton Handlirsch. In nine parts, pp. 1430, 
many plates and text-figures. Leipzig, 1906-08 (Wm, Engel- 
mann). — This monumental work reviewing 7640 species of fossil 
insects is now completed. (An earlier review appeared in April, 
1907, of this Journal.) On pages 1142-1 192 is given a summary of 
the paleontologic results, a chapter of the greatest interest not only 

* It is proposed that the magnetic charts of the Carnegie Institution, with 
which Dr. Bauer is now connected, be also referred to the same period. 



2CA Scientific Intelligence. 

to entomologists but as well to all paleontologists. The first 
insects found fossil occur at the base of the Pennsylvanian or 
Lower Coal Measures. Of Paleozoic forms there are 884 species,. 
Mesozoic 965 and of Cenozoic 5802. Rather a small showing 
when compared with the 380,000 known recent species, and yet a 
number sufficiently large to give a clear insight into the chrono- 
genesis of the winged insects. 

In Chapter VII a review is given of the more important classi- 
fications of recent insects beginning with that of Aristotle 300 
years before Christ. In the eighth chapter the author establishes 
a new classification based on the principles of phylogeny and 
chronogenesis. There are two classes of insects, the winged or 
Pterygogonea, and the wingless or Apterygogonea. The former 
is here divided into 12 subclasses and 39 orders. The oldest sub- 
class, restricted to the Pennsylvanian, is the radical stock Palse- 
odictyoptera and out of it are derived 11 primary orders, all but 
one of these being restricted to the Paleozoic. The exception is 
Blattoidea. Either late in the Permian or in the Triassic origi- 
nate 16 terminal orders, followed by 3 in the Jurassic, 7 in the 
Cretaceous and 2 in the Cenozoic. 

The phylogeny of the Arthropoda is discussed on pages 1293— 
1318. The Trilobita is regarded as the radical from which most 
of the Arthropoda phyla? are directly derived. Finally, on pages 
1318-1344, the author presents his "descendance conclusions."" 
A Nobel prize might well be the reward for the energy and 
insight here displayed. c. s. 

2. Connecticut Geological and Natural History Survey. Third 
Biennial Report ; Bulletin No. 12, 1907-1908. William North 
Rice, Director. 30 pages. Hartford. — The report of the Survey 
shows an unusual amount of high grade work accomplished with 
a very small annual appropriation. In- addition to the eleven 
bulletins already published, the following have been accepted : 
The Lithology of Connecticut ; the Flowering Plants and Pteri- 
dophytes of Connecticut ; the Hymeniales of Connecticut ; the 
Triassic Fishes of Connecticut ; the Insects of Connecticut. The 
work in progress includes researches in glacial geology, study of 
peat deposits, and the preparation of reports on fresh waters, pro- 
tozoa and algae, echinodermata, and birds of Connecticut. 

H. E. G. 

3. Mississippi State Geological Survey. Albert F. Crider, 
Director. Bulletin No. 1, Cement and Portland Cement Mate- 
rials of Mississippi. 70 pp., 6 pis. Bulletin No. 2, Clays of Mis- 
sissippi, pt. 1, pp. 249, pis. 42, figs. 14. Bulletin No. 3, The 
Lignite of Mississippi, pp. 66. Bulletin 4, Clays of Mississippi, 
pt. ii, pp. 64, pis. 17. — The Mississippi survey is justifying its 
existence from an economic standpoint by giving its first atten- 
tion to the resources of the state, as shown by the above titles. 
Bound with each bulletin is a copy of the new Geologic and 
Topographic Map, which " represents the present knowledge of 
geologic boundaries of Mississippi " — a publication which will be 
welcomed by geological students. h. e. g. 



Geology and Natural History. 265 

4. The Interpretation of Topographic Maps ; by Rollin D. 
Salisbury and Wallace W. Atwood. U. S. Geological Sur- 
vey, Professional Paper No. GO. Pp. 78, pis. 170, figs. 34. 
Washington, 1908. — The use of the topographic maps published 
"by the United States Geological Survey will doubtless be greatly 
extended as the result of this publication. The interpretation of 
156 maps is given and they are grouped, with reference to the 
features the}' represent, under the following heads : The work of 
wind ; stream erosion ; alluviation ; topographic effects from 
unequal hardness of rocks ; erosion cycles ; stream piracy and 
adjustment ; effects of ground water ; glaciation ; coast lines ; 
volcanism; faults; lakes of special types. From the standpoint 
of the teacher, this paper is one of the most valuable ever issued 
by the Geological Survey. Its value could have been increased, 
however, by more complete explanations of the various topo- 
graphic types. h. e. g. 

5. The Zonal Belt Hypothesis, a New Explanation of the 
Causes of the Ice Age; by Joseph T. Wheeler. Pp. 401. 
Philadelphia and London, 1908 (J. B. Lippincott Co.). — An 
attempt is made, in this book, to account for the great changes 
of climate through geological time, on the assumption that the 
earth has been, until recent times, surrounded by belts of 
"planetesimal or gaseous matter." The author has apparently 
read widely, in search of geological and meteorological pheno- 
mena which might be explained by this hypothesis. On the 
theory that the last zonal belt was in existence during the early 
history of man, the myths of the various races, together with 
Genesis and Plato, are called in as proofs of this phenomenon. 
Although the book contains much interesting matter, it can 
hardly be classed as a scientific treatise. h. e. g. 

6. Handbuch, cler Miner alo g ie ; von Carl Hintze. Erster 
Band. Zwolfte Lieferung. Pp. 1761-1920; 43 text-figures. 
Leipzig 1908 (Veit & Co.). — The twelfth part of the first 
volume of Hintze's great work on Mineralogy has recently been 
issued. This is largely devoted to exhaustive descriptions of the 
species hematite and ilmenite ; also of several of the protoxides, 
including cuprite. This part is the twenty-fourth of the entire 
work begun in 1889; a title page is issued with it to be used in 
binding the first part of Vol. I (pp. 1-1208), embracing the 
elements and sulphides. h. e. g. 

7. Chemische Krystallographie ; von P. Groth. Zweiter 
Teil. Die anorganischen Oxo- und Sulfosalze. Pp. 914, 522 
text-figures. Leipzig, 1908 (W. Engelmann). — The scope of the 
important work on Chemical Crystallography undertaken by 
Prof. Groth was discussed in a notice of the first part published 
in 1906 (volume xxiii, p. 153). We have now the second 
instalment of the work, a volume of more than 900 pages, 
embracing the Inorganic Oxy- and Sulphosalts. This volume 
will be of especial interest to mineralogists, as well as to chemists, 
since it contains so large a part of the well-known mineral com- 



266 Scientific Intelligence. 

pounds, as well as the allied artificial salts. The descriptions are 
liberally illustrated, and the statement of the crystallographic 
data leaves nothing to be desired. The work, however, is much 
more than a compilation of the crystallographic and optical prop- 
erties of chemical compounds, for the author's breadth of view 
and philosophical insight into the relations of these compounds 
give great value to the introductory remarks w T ith which each 
division is introduced. The friends of the author will wonder 
once more that he is able, in connection with so many, other 
works of importance, to carry through so stupendous a labor as is 
here represented. Two more parts are in preparation, which will 
embrace the organic compounds. 

8. Notes of a Botanist on the Amazon and Andes, being 
records of travel on the Amazon and its tributaries; as also to 
the cataracts of the Orinoco, along the eastern side of the Andes 
of Peru and Ecuador, and the shores of the Pacific, during the 
years 18^9-186^ ; by Richard Spruce, Ph.D. Edited and 
condensed by Alfred Russel Wallace, O.M., F.R.S., with a 
biographical introduction, portrait, seventy-one illustrations, and 
seven maps. In two volumes. London, 1908 (Macinillan & Co).— 
This treatise is exceptional in many respects. It is the record of 
botanical exploration about half a century ago, which is almost 
as fresh and important as if it was made during the last year. 
Furthermore, it is full of physio^raphical and ethnological 
memoranda of extraordinary interest, edited by a thoroughly 
sympathetic friend and fellow-naturalist. Everybody knows the 
range of the journeys made by Wallace in South America, and is 
familiar with the hardships attending them. Spruce not only 
passed over a great deal of territory which was practically like 
that investigated by Wallace, but he was in the equatorial belt 
at the very time when Wallace was there and knew of his serious 
illness. 

Spruce is best known for his copious contributions to our 
knowledge of the Bryophytes. Beginning their study during his 
life in Yorkshire, he carried on explorations in the Pyrenees, 
where he discovered mosses and liverworts in places which had 
been thought to contain none, and afterwards he collected exten- 
sively in South America. With scarcely any means, and with 
only enfeebled health, he managed, by dint of an iron will, to 
carry his work on in a manner which has -always commanded 
respect. 

He has been most fortunate in his editor. No onejcould have 
condensed the voluminous notes more skillfully, or connected 
them with more instructive remarks than Mr. Wallace. 

Our readers will enjoy in the perusal of this absorbingly inter- 
esting volume the proofs of Spruce's sagacity which led him to 
interpret many structural features in the Tropics as indicative of 
a certain drifting of specific characters, vaguely pointing towards 
descent through variation. After Spruce's return to England, he 
became an ardent Darwinian, going so far as to state unequivo- 



Miscellaneous Intelligence. 267 

cally that "if we had all the forms now in existence, arid that 
have ever existed, of such genera as Rubus, Asplenium, Bryum, 
and Plagiochila, we should be unable to define a single species — 
the attempt to do so would only be trying to separate what Na- 
ture never put asunder — but we should see distinctly how certain 
peculiarities had originated and become temporarily fixed by 
inheritance ; and we could trace the unbroken pedigree of every 
form. " G. l. g. 

III. Miscellaneous Scientific Intelligence. 

1. Carnegie Institution of Washington. Year Book No. 7, 
1908. Washington, 1909. — The following is an authorized state- 
ment of the contents of the seventh Year Book of the Carnegie 
Institute of Washington, for 1908 : It consists of reports of the 
President and the Executive Committee, and of Directors of 
Departments and other grantees who, with the assistance of the 
Institution, have been carrying on investigations during the year. 

The President's report gives the following salient facts and 
figures indicating the growth and extent of the work thus far 
undertaken and accomplished by the Institution. Since its organ- 
ization, in 1902, about 1,000 individuals have been engaged in 
investigations under the auspices of the Institution and there are 
at present nearly 500 so engaged. Ten independent departments, 
each with its staff of investigators and assistants, have been 
established. In addition to these larger departments of work, 
organized by the Institution itself, numerous special researches, 
carried on by individuals, have been subsidized. Six laboratories, 
for as many different fields of investigation, and in widely sep- 
arated localities, have been constructed and equipped. Work in 
almost every field of research, from archeology and astronomy 
to thermodynamics and zoology, has been undertaken, and the 
geographical range of this work has extended to more than thirty 
different countries. 

At the end of the fiscal year, October 31, 1908, 120 volumes of 
researches in nineteen different fields of research, with an aggre- 
gate of more than 30,000 pages, had been published, and 27 
volumes of research were in press. In addition to these publica- 
tions issued by the Institution, about 1,000 shorter papers have 
been published in the current journals of -the world by depart- 
ment investigators, by associates, and by assistants. The total 
amount of funds appropriated for expenditure to November 1 
was $3,683,840.00, which included $293,928.37 reverted and 
afterward reappropriated. The total amount expended was 
$3,359,236.17. 

During the past year the Nutrition Laboratory in Boston has 
been equipped, and systematic investigations are already in 
progress. 

The construction of a building in Washington, D. C, at the 
corner of Sixteenth and P streets, N. W., was begun a year ago. 
This building is for administrative offices and the storage of 



268 Scientific Intelligence. 

records and publications, and when completed will cost about 
$220,000. 

The plans and specifications for the construction of a specially 
designed ship for ocean magnetic work have recently been com- 
pleted. These plans call for a non-magnetic sailing vessel with 
auxiliary propulsion, and the contract for her construction has 
been let to the Tebo Yacht Basin Company, of Brooklyn, N. Y. 
She will be classified as a yacht, will be called the " Carnegie," 
and will, upon completion, proceed upon a magnetic survey of 
the Atlantic Ocean under the direction of the Department of 
Terrestrial Magnetism of the Institution. 

A temporary observatory for supplementary measures of the 
positions of the fixed stars of the southern hemisphere is now being- 
built at San Luis, Argentina, under the direction of Professor 
Lewis Boss, head of the Department of Meridian Astrometry of the 
Institution. Professor R. H. Tucker will be resident astronomer in 
charge of the work of observing and computing in South Amer- 
ica, which will require three to five years for completion. The 
meridian instrument of the Dudley Observatory, whose constants 
have been thoroughly investigated, will be transferred to San 
Luis and used in securing the desired measurements of the posi- 
tions of stars in both hemispheres. 

Work in the other departments of the Institution has pro 
gressed rapidly and successfully. The investigations of Dr. G. 
E. Hale, Director of the Solar Observatory on Mount Wilson, 
California, are of great interest. During the .year, with the aid 
of his exceptional equipment, certain discoveries with regard to 
sun-spots have been made which will probably prove of as great 
importance to terrestrial and molecular physics as to solar physics. 
The progress inaugurated may be confidently expected to lead 
rapidly to definite and important results. Under the direction of 
the Department of Historical Research, work upon manuscript 
materials for American History has been pursued in France, 
Italy and England, and next year will be extended to Germany. 
Many remarkable experiments and investigations are in progress 
under the Department of Botanical Research at the Desert Lab- 
oratory at Tucson, Arizona. 

In addition to the work carried on in the departments of the 
Institution during the year, 31 grants were made to individuals 
and organizations, in aid of researches conducted by them, and 
many other researches begun in former years have been carried 
forward. The publication of 20 volumes was authorized, and 27 
volumes and an atlas have been published. These latter include 
the report upon the California Earthquake of April 18, 1906, a 
Handbook of Learned Societies and Institutions of North and 
South America, and a reproduction of the "Old Yellow Book," 
the source of Browning's u The Ring and the Book." These 
volumes and others issued by the Institution are offered for sale 
at the cost of printing and transportation to purchasers. 

At the annual meeting of the Board of Trustees on December 
8, 1908, Mr. Martin A. Ryerson, of Chicago, was elected a 



Miseellan eous In tettigcn ce. 269 

Trustee, to fill a vacancy in the Board. At the same meeting 
the sum of 8636,300 was appropriated to carry on work of inves- 
tigation, publication and administration during the year 1909. 

2. Report of the Librarian of Congress and Report of the 
Superintendent of the Library Buildings and Grounds for the 

fiscal year ending June 30, 1908. Pp. 143. Washington, 1908. 
— The Library of Congress is so far a model in its work and 
methods for the other large libraries in the country that the 
annual report of the Librarian, Mr. Herbert Putnam, possesses 
particular interest. The total amount expended by the Library, 
in 1908, was -$615,000, while the appropriations for 1909 are 
somewhat more than $100,000 greater than this sum. The 
increase in number of books for the year is slightly over 100,000. 
The important purchase of the year was that of the Huitfeldt- 
Kaas collection of 5000 volumes of Scandinavian literature, made 
by the state archivist of Norway, who died in 1905. A portrait 
of Mr. A. R. Spofford, who was for 32 years Librarian in Chief, 
and who died on August 11, forms the frontispiece of the volume. 

3. Harvard College Observatory ; Edward C. Pickering, 
Director. — Recent publication are noted in the following list 
(continued from vol. xxvi, pp. 99): 

Annals. — Vol. LIV\ A Catalogue of Stars fainter than the 
magnitude 650 observed with the 4-inch Meridian Photometer; 
forming a supplement to the Revised Harvard Photometry. 
Pp. 280. 

Vol. LVI, No. IV. Classification of 1,477 Stars by means of 
their Photographic Spectra; by Annie J. Cannon. Pp. 65-114. 

Vol. LVII, Part II, Comparison Stars for 252 Variables of 
Long Period. Prepared for publication by Leon Campbell, 
Assistant, under the direction of Edward C. Pickering, Director. 
Pp. 213-288. 

Vol. LIX, No. II. Photographic Photometry on a Uniform 
Scale ; by Edward S. King. ^Pp. 33-62/6 figures. 

Vol. LX, No. IX. A Catalogue of Photographic Charts of 
the Sky. Pp. 231-251. 

Vol. LXlV, No. I. Observations with the Meridian Photo- 
meter during the years 1902 to 1906. Pp. 32. No. II. The 
Variable Star SS Cygni. 213843 ; by Leon Campbell. Pp. 33-53. 
I Plate. No. III. Schonfeld's Comparison Stars for Variables. 
Pp. 55-89. 

Circulars. — No. 137. 25 New Variable Stars in Harvard 
Map, Nos. 31-43. 

No. 138. 060547. The Variable Star, 31,1907. Pp. 6. 

No. 139. 26° 179. A New Variable of the Class of Lyra?. 
003226. Pp. 7. 

No. 140. 16 New Variable Stars in Harvard Map, Nos. 4 and 
13. Pp. 3. 

No. 141. 29 New Variable Stars near Nova Sagittarii. Pp. 4. 

No. 142. 28 New Variable Stars in Harvard Map, Nos. 30 and 
33. Pp. 3. 



270 Scientific Intelligence. 

4. Publications of the Allegheny Observatory of the Uni- 
versity of Pittsburgh. — The following have recently appeared : 

Vol. I, No. 6. The determination of the orbit of a spectro- 
scopic binary by the method of least squares ; by Frank Schles- 
inger. Pp. 33-44. 

No. 7. The orbit of Aquilse ; by Robert H. Baker. Pp. 
45-66. 

No. 9. A partly graphical method for predicting solar eclipses; 
by Frank Schlesinger. Pp. 57-64. 

5. Washburn Observatory of the University of Wisconsin, 
George C. Comstock, Director. — The following volume has been 
recently issued : 

Vol/ XII. Determinations of Proper Motion, 1902-1907. 
Part I. Proper Motions of Faint Stars ; by George C. Comstock, 
Director. Pp. 317. The observations here given have been 
made with the 40 cm. equatorial of the Washburn Observatory; 
to a considerable extent they repeat, after a period of more than 
"fifty years, the measurements made by Struve at Pulkova, on stars 
fainter than the ninth magnitude. 

6. A Treatise on Spherical Astronomy • by Sir Robert Ball. 
Cambridge, 1908 (The University Press). — The present treatise 
by Sir Robert Ball, the distinguished Lowndean Professor of 
Astronomy at Cambridge University, covers quite an amount of 
ground not included in the classical works of Brtinnow, Chauve- 
net and later writers and thus seems a valuable addition to our 
text-books. Thus it gives a glimpse into the theory of map 
making and into the rudiments of theoretical astronomy besides 
the usual applications of spherical trigonometry to celestial prob- 
lems. An interesting feature are the numerous exercises, though 
some of these seem more a test of algebraic ingenuity than of 
astronomical insight. Among the novelties of the book may be 
cited the nomenclature "nole" and " anti-nole " for the poles 
of a great circle, corresponding to the North and South Poles ; 
to us it hardly seems that there was a crying need for further 
designations, also, perhaps, the similarity to " node " is unfor- 
tunate. 

The presentation, as is usual with Sir Robert Ball, combines 
great lucidity and directness and we may again add our best 
wishes for the adoption of the work for class and private study. 

w. L. E. 

7. Bulletin of the Mount Weather Observatory, William J. 
Humphkeys, Director. — Prepared under the direction of Willis 
L. Moore, Chief U. S. Weather Bureau. Vol. I. Part 4, pp. 
207-77 ; 3 charts, 3 figures. — This number contains the following 
articles : Pyrheliometer and polarimeter observations, by H. H. 
Kimball ; recent auroral displays and magnetic disturbances, by 
W. R. Gregg ; magnetic declination, by E. R. Miller ; upper air 
temperatures for April, May and June, with charts of upper air 
isotherms, by W. R. Blair. 



Miscellaneous Intelligence. 271 

8. National Antarctic Expedition, 190 1-190 L£, Album of 
Photographs and Sketches with a Portfolio of Panoramic Views. 
Pp. 303, 4to ; 165 plates and 2 maps. London, 1908. (Pub- 
lished by the Royal Society and sold by Harrison & Sons, St. 
Martin's Lane, W. C, and Oliver & Boyd, Tweeddale Court, 
Edinburgh.) — The two volumes giving the results of the meteor- 
ological and physical observations made by the National 
Antarctic Expedition were noticed in the December number 
(p. 588). We have now an additional volume presenting in most 
attractive form views of the life and scenery of the region visited. 
The photographic work was in charge of Lt. Skelton, and most 
of the admirable views here reproduced were taken by him ; the 
name of L. C. Bernacchi is also frequently associated with his. 
The volume and accompanying portfolio contain 165 plates, of 
w r hich 128 are from photographs. The life of the southern ocean 
is well shown in the interesting series of photographs of the 
Adelie and Emperor penguins, also of the seals, whales and 
albatross. The scenic views present a vivid impression of the 
striking features of the land and ice of the Antarctic, particularly 
of its capes and lofty mountains. Especial interest attaches to the 
photographs of the Great Ice Barrier, the abrupt cliffs of which ex- 
tend for many miles from King Edward VII Land to Ross island. 
The height is as great as 280 feet at some points, and the surface 
comes down at others to within 18 i'eet of the water level. This 
Great Barrier sheet moves northward at the rate of 45 yards per 
month, but in consequence of its recession by breaking up the 
sea face is now some 10 to 15 miles farther south than sixty years 
ago. A special portfolio contains an interesting series of sketches, 
largely panoramic, due to the artistic work of the Junior Sur- 
geon, Edward A. Wilson. 

9. The Chemical Constitution of the Proteins; by R. H. 
Aders Plimjier, D.Sc. In two parts. Part I, pp. xii + 100; 
Part II, pp. xii-f 66. London and New York, 1908 (Longmans, 
Green and Co.). — The development of biochemistry as a sepa- 
rate department of study has been so rapid and fruitful that the 
entire field is already too large for exhaustive treatment in a 
single volume. The unique importance of the proteins has 
always made them favorite subjects for investigation. In recent 
years particularly, the literature on these compounds has increased 
to an extent that makes timely an exhaustive compilation like 
Dr. Plimmer's. The monographs are concerned with the chem- 
ical composition of the protein molecule and the chemical charac- 
teristics of its component units, rather than with the behavior of 
individual protein substances. The organic chemist as well as 
the physiologist will have occasion to consult such a review, in 
considering the historic development of the study of the amino- 
acids as well as the interesting attempts at the synthesis of com- 
plex protein-like compounds (with which Part II especially 
deals). In accord with the intent of this series of monographs, 
an extensive bibliography is published in each part. l. b. m. 



272 Scientific Intelligence. 

10. Standard Algebra ; by William J. Milne, New York 
State Normal College, Albany, N. Y. Half leather, 12 mo, 464 
pages. New York, 1908 (American Book Company). — This use- 
ful text-book follows the method of the Inductive Algebra pub- 
lished in 1881 by the same author. The new ideas that have been 
brought into the teaching of elementary algebra in the interval 
are presented, such as the remainder theorem for purposes of 
factoring, and the graph, which is used for illustration to a limited 
degree. It is a handy volume, bound and trimmed to carry 
in the pocket, and the exercises are numerous, fresh and well 
chosen. w. b. 

11. Kraft, 6ko7iomisehe, technische unci kulturgeschichtliche 
St'iidien uber die Machtentfaltung der Staaten / von E. Reykr. 
8°, pp. 380, many diagrams in text. Leipzig, 1908. — The work of 
Professor Reyer, who is well known for his studies in theoretical 
geology, has always been characterized by great originality, both 
in his views and in his mode of presenting them. He has also 
been widely interested in matters humanistic, and it is, there- 
fore, not strange to find a volume coming from his pen which 
relates to subjects outside of his especial held. 

The volume before us is a conspectus of the economic develop- 
ment of the chief nations in matters relating to their main indus- 
tries, those upon which their -material prosperity depends. Such 
comprise iron and steel, coal, precious metals, manufactures,, 
commerce on land and sea, agriculture, and a great variety of 
others in which power is an element of first importance. The 
history of the growth of power, its various sources of supply, its 
relation to the topics mentioned above, and its future are thought- 
fully treated. The book is illustrated by many diagrams and, 
written in a clear and well-poised manner, it presents an interest- 
ing and popular exposition of a most important subject.. 

l. v. p. 

12. International Congress of Applied Chemistry. — The seventh 
International Congress of Applied Chemistry will be held 
in London from May 2V to June 2. The Honorary Secretary is 
William Macnab, address 10 Cromwell Crescent, S.W. 

The Science Year Book, with Astronomical, Physical and Chemical 
Tables ; Summary of Progress in Science, Directory, Biographies and Diary, 
for 1909 (fifth year). Edited by Major B. F. S. Baden-Powell. Portrait of 
Sir William Ramsay as frontispiece. Pp. 148 + 365. London (King, Sell 
& Olding, 27 Chancery Lane, W.C.) 

Obituary. 
Professor Harry G. Seeley, the well-known English geolo- 
gist and vertebrate paleontologist, died on January 8, in his 
seventieth year. 



New Circulars, 



84 : Eighth Mineral List : A descriptive list of new arrivals, 

rare and showy minerals. 

85 : Minerals for Sale by Weight: Price list of minerals for 

blowpipe and laboratory work. 

86: Minerals and Rocks for Working Collections: List of 
common minerals and rocks for study specimens ; prices 
from i^ cents up. 

Catalogue 26: Biological Supplies : New illustrated price list 
of material for dissection ; study and display specimens; 
special dissections; models, etc. Sixth edition. 

Any or all of the above lists will he sent free on request. We are 
constantly acquiring new material and publishing new lists. It pays to 
be on our mailing list. 

Ward's Natural Science Establishment 

76-104 College Ave., Rochester, N. Y. 



Ward's Natural Science Establishment 

A Supply-House for Scientific Material. 

Founded 1862. Incorporated 1890. 

DEPARTMENTS : 

Geology, including Phenomenal and Physiographic. 
Mineralogy, including also Rocks, Meteorites, etc. 
Palaeontology. Archaeology and Ethnology. 

Invertebrates, including Biology, Conchology, etc. 
Zoology, including Osteology and Taxidermy. 
Human Anatomy, including Craniology, Odontology, etc. 
Models, Plaster Casts and Wall-Charts in all departments. 



Circulars in any department free on request ; address 

Ward's Natural Science Establishment, 

76-104 College Ave., Rochester, New York, IT. S. A. 



CONTENTS. 

Page 
Art. XII. — Recent Observations in Atmospheric Electricity; 

by P. H. Dike _ 197 

XIII. — Iodyrite from Tonopah, Nevada, and Broken Hill, 

New South Wales ; by E. H. Kraus and C. W. Cook__ 210 

XIV. — Deviation of Rays by Prisms; by H. S. Uhler 223 

XV. — Heat of Oxidation of Tin, and second paper on the 
' Heat of Combination of Acidic Oxides with Sodium 
Oxides ; by W. G. Mixter 229 

XVI. — Neptunite Crystals from San Benito County, Cali- 
fornia ; by W. E. Ford ._ _. 235 

XVII. — Gravimetric Determination of Silver as the Chro- 

mate ; by F. A. Gooch and R. S. Bosworth 241 

XVIII. — Doppler Effect in Positive Rays; by J. Trowbridge 245 

XIX. — New Armored Saurian from the Niobrara ; by G. R. 

Wieland _ _ 250 

WOLCOTT GlBBS _ _ _ 253 

SCIENTIFIC INTELLIGENCE. 

Chemistry and Physics— New Method of Forming Liquid Alloys of Sodium 
and Potassium, Jaubert : Determination of Cerium and other Eare Earths 
in Eocks, Dietrich, 260. — Solvability of Metallic Gold in Hydrochloric 
Acid in the presence of Organic Substances, Awerkiew : The Composi- 
tion of Matter, E. Mulder, 261. — Introduction to the Earer Elements, P. 
E. Browning : Feste Losungen und Isomorphism, G. Bruni : Produc- 
tion of Helium from Uranium, F. Soddy : Charge and Nature of the 
a-Particle, E. Eutherpord and H. Geiger : Amount of Water in a Cloud 
formed by Expansion, W. B. Morton, 262. — Permanent Magnetism of 
Copper, J. G. Gray and A. D. Eoss : United States Magnetic Tables and 
Magnetic Charts for 1905, L. A. Bauer, 263. 

Geology and Natural History— Fossilen Insekten und die Phylogenie der 
Eezenten Formen, A. Handlirsch, 263. — Connecticut Geological and 
Natural History Survey, W. N. Eice : Mississippi State Geological Sur- 
vey, A. F. Crider, 264. — Interpretation of Topographic Maps, E. D. 
Salisbury and W. W. Atwood : Zonal Belt Hypothesis, a New Explana- 
tion of the Causes of the Ice Age, J. T. Wheeler : Handbuch der Min- 
eralogie, C. Hintze : Chemische Krystallographie, P. Groth, 265. — Notes 
of a Botanist on the Amazon and Andes, E. Spruce, 266. 

Miscellaneous Scientific Intelligence — Carnegie Institution of Washington. 
Year Book No. 7, 1908, 267.— Eeport of the Librarian of Congress and 
Eeport of the Superintendent of the Library Buildings and Grounds, H. 
Putnam : Harvard College Observatory, E. C. Pickering, 269. — Publica- 
tions of the Allegheny Observatory of the University of Pittsburgh : 
Washburn Observatory of the University of Wisconsin, G. C. Comstock : 
Treatise on Spherical Astronomy, E. Ball : Bulletin of the Mount 
Weather Observatory, W. J. Humphreys, 270. — National Antarctic Expe- 
dition, 1901-1904 : Chemical Constitution of the Proteins, E. H. Aders 
Plimmer, 271. — Standard Algebra, W. J. Milne: Kraft, okonomische, 
technische und kulturgeschichtliche Studien uber die Machtentfaltung der 
Staaten, E. Beyer : International Congress of Applied Chemistry : The 
Science Year Book, 272. 

Obituary— H. G. Seeley, 272. 



Dr. Cyrus Adler, 

Librarian U. S. Nat. Museum. 



YOL. XXVII. 



APRIL, 1909. 



Established by BENJAMIN SIILIMAN in 1818. 



THE 



AMERICAN 

JOURNAL OF SCIENCE 



Editor: EDWAED S. DANA. 



ASSOCIATE EDITORS 



Professors GEORGE L. GOODALE, JOHN TROWBRIDGE, 
W. G. FARLOW and WM. M. DAVIS, of Cambridge, 

Professors ADDISON E. VERRILL, HORACE L. WELLS, 
L. V. PIRSSON and H. E. GREGORY, of New Haven, 

Professor GEORGE F. BARKER, of Philadelphia, 
Professor HENRY S. WILLIAMS, of Ithaca, 
Professor JOSEPH S. AMES, of Baltimore, 
Mr. J. S. DILLER, of Washington. 



FOURTH SERIES 

No. 160— APRIL, 1909. 

VOL. XXVII-[WHOLE NUMBER, CLXXVIL] 



NEW HAVEN, CONNECTIC 

1 909. 




THE TUTTLE, MOREHOUSE & TAYLOR CO., PRINTERS, 123 TEMPLE STREET. 



Published monthly. Six dollars per year, in advance. $6.40 to countries in the 
Postal Union ; $6.25 to Canada. Keinittances should be made either by money orders 
registered letters, or bank checks (preferably on New York banks) 



BAKE CINNABARS FKOM CHINA. 

We would call attention to these remarkable Cinnabars, some of which 
still remain. They are beyond doubt the most beautiful and interesting 
crystals of Cinnabar ever discovered. They were described and illustrated 
in this Journal, Nov. '08. The prices range from $5, $7.50, $10, $15, $18, 
$25, $35, and $50. 

NEW ABRIVALS.— RARE FINLAND MINERALS. 

Wiikit, Idocrase, Gigantolite, Tantalite, Skapolite, Sordavalite, Pitkaran- 
dite, Columbite, Monazite, Magnetite, Pyroxene, Sundvikite, Amphibole, 
Chrysoberyl, Muscovite curved, Epidote, Cassiterite with Pyrite, Beryls, 
Garnets, Tourmaline, Pyrites, Chondrodite, Quartz, Rose Quartz. 

EAEE MINERALS— EEOSSNITZ ALPS, TYROL. 

Bornite, largest crystal found, 1£ by 1 in., slightly altered. Bornite, 
altered to malachite, with gold, loose and in matrix. Brookite, flat loose 
crystals, rare form ; Eisenrose, large rare crystals ; Anatase, small loose 
crystals and in matrix ; Smoky quartz with Byssolite ; Smoky Quartz crys- 
tals, Habachthal ; Albite, choice rare crystals ; Sphene, Habachthal ; Apa- 
tites, in matrix, beautiful white, Untersulzbachthal ; Emeralds doubly 
terminated in matrix, Habachthal ; Epidote, large gem crystal, 5 in. long. 

RARE MINERALS FROM OTHER LOCALITIES. 

Euclase, Capo do Lane, Brazil; Columbite, Conn.; Monazite, large loose 
xls. and in matrix, Conn.; Uraninite, crystal in matrix, Conn.; Benitoite 
and Neptunite, San Benito Co. , Cal.; Lievrite, Elba; Polybasite, Hungary 
and Mexico ; Herderite, Poland, Maine ; Cobaltite, loose xls. and in matrix, 
Cobalt, Ont. , and Sweden ; Yivianite, large crystals, Colo. ; Olivinite, Utah ; 
Sartorite, Canton Wallis ; Jordanite, Binnenthal ; Crocoite, Siberia and Tas- 
mania ; Cinnabar, China, Cal., and Hung.; Gypsum, twin xls., Thuringia ; 
Diamond in matrix, New Yaal River Mine, South Africa ; Argentite, Mex. ; 
Freiberg, Sax.; Pyrargyrite, Sax. and Mex.; Pyromorphite, Ems, Germany, 
Phoenixville, Pa. ; Tourmalines, beautiful sections from Brazil, Mesa Grande, 
Cal., all shades, in matrix and loose xls.; Pink Beryl, small and large, Mesa 
Grande, Cal.; Kunzite, small and large, Pala, Cal.; Titiantite, Tilly Foster; 
N. Y.; Tetrahedrite, Utah and Hungary ; Realgar, Hung.; Opal, Caribou River, 
Queensland ; Bismuth, native, Cobalt, Ont.; Zircon xls., loose, Ural ; Green 
and Cinnamon Garnets, Minot, Maine ; Yesuvianite, Poland, Maine, Italy 
and Tyrol ; Zeolites, beautiful specimens from Erie Tunnel, Patterson, and. 
Great Notch. 

Reconstructed Rubies, Blue and White Sapphires. 

Do not fail to secure one of these beautiful scientific gems ; the blue and 
white sapphires are the latest discovery ; they have the same specific gravity 
and hardness as the genuine ; the brilliancy and color equal to the genuine 
and better. 

We have every known cut precious and semi-precious stones. 

Let us know your wants, and we will send the specimens on approval to 
you. Further particulars cheerfully furnished. 

A. H. PETEREIT, 

81—83 Fulton Street, New York City. 



1 



THE 



AMERICAN JOURNAL OF SCIENCE 

[FOURTH SERIES.] 



Art. XX.— On the Permeabilities and the Reluctivities, for 
very Wide Ranges of Excitation, of Normal Specimens of 
Compressed Steel, Bessemer Steel and JVorivay Iron Rods ; 
by B. Osgood Peikoe. 

When a rod or a closed frame of iron becomes magnetized 
under the action of a steady electric current in an exciting 
coil of insulated wire wound about it, the flux of magnetic 
induction (B) through any cross section of the iron can be 
easily determined with the' aid of a small testing coil, but it is 
often very difficult to tell just what the value of the exciting 
magnetic' field (B) is at any given point within the metal. In 
a few familiar cases, however, the difficulty disappears. 

If, for instance, a homogeneous round rod of soft iron the 
length of which is,. say, hVe hundred times the diameter, be 
placed in a solenoid of narrow bore, somewhat longer than the 
rod and uniformly wound with n turns of insulated wire per 
centimeter of its length, and if a steady current (G) be sent 
through the solenoid, the demagnetizing effects of the ends of 
the rod become inappreciable near the center, O, and we 
may assume without sensible error that the value of B at every 
point of a cross section in the neighborhood of O is equal to 
the value (^irnC/10) of // just without the rod at O. 

In the case of a soft iron toroid, uniformly wound with JV 
turns (in all) of insulated wire, the value of Bis, of course, 
not the same at every point of a meridional section of the 
metal, but if the material is perfectly homogeneous, there is 
practically no leakage of induction into the air. Under 
these circumstances, the magnetomotive force is the same 
(■i-n-jVC/lO) around all closed non-evanescible paths in the 
iron, and the value of B is inversely proportional to the dis- 
tance from the axis of revolution of the toroid. In practice, 
kit Jour. Sci.— Fourth Series, Vol. XXVII, No. 160.— April, 1909. 
19 



274 Peirce — Permeabilities and Reluctivities for Steel. 

such a toroid must be turned out of a solid piece of the metal, 
for it is almost impossible to make a ring out of iron rod, bent 
and welded together at the ends, which shall be sufficiently 
homogeneous to prevent serious leakage of lines of induction, 
and IT is often very different in such a ring at different points 
of a circumference in the iron coaxial with the ring. If a closed 
magnetic circuit be made by putting a piece of soft iron rod 
lengthwise between the jaws of a massive yoke of any of the 
common forms, there is usually such an amount of leakage 
through the surface of the rod that the induction flux has very 
different values at different sections of it. However carefully 
the joints are made between the jaws and the specimen, it is 
very difficult to compute, from previous determinations of the 
magnetic properties of the rod and the yoke, what portion of 
the whole magnetomotive force of the circuit is used in the 
rod ; indeed the fraction is very different for different excita- 
tions, and for the same soft yoke may depend very much upon 
the hardness of the piece to be studied. For comparatively 
low excitations up to say H= 100, a slender yoke may be used 
so that the cross section of the magnetic circuit is not very 
different at different places, and the exciting coils can then be 
so arranged on the yoke and the specimen, if the joints are 
well made and the whole circuit is magnetically fairly homo- 
geneous, that the induction flux is nearly the same throughout. 
It is understood that this procedure* has been brought to 
great perfection in the National Bureau of Standards at Wash- 
ington. For excitations of //=2000 or more, however, the 
arrangement seems hard to manage. 

A magnetic field in air near a magnet is both solenoidal and 
lamellar, and if in any portion of such a field the lines are 
sensibly straight and parallel, we may infer that in that region 
the field is practically uniform. If within a piece of perfectly 
soft iron the magnetization vector (/) has everywhere the 
direction of the exciting magnetic force, IT, and an intensity 
expressible in terms of H alone, the induction in the metal is 
a solenoidal vector which has the same lines as the lamellar 
vector H. If throughout any region within such a piece of 
iron the lines of force are straight and parallel, the magnetic 
force and the induction are both uniform in that region. At 
the surface of separation between iron and air, the tangential 
components of the force are continuous, but the normal com- 
ponent of the force is generally discontinous ; if the lines of 
force in air just outside the surface are parallel to it, the lines 
of induction in the metal at this surface are parallel to the 
surface. If, then, the magnetic field around a slender rod of 
* Burrows, Proc. American Soc. for Testing Materials, viii, 1908. 



Peirce — Permeabilities and Reluctivities for Steel. 275 

iron magnetized lengthwise, has, near the surface of the rod, 
lines parallel to the rod's axis, we may inquire whether the 
lines of force and of induction within the rod are not in this 
neighborhood all parallel to the axis, so that the value of H 
throughout a cross section of the rod is the same as the value 
just outside the iron. 

The assumption that the value of PL in the air just about a 
slender neck of iron of proper form held between the jaws of 
a highly excited electromagnet is the same as the value in the 
neck itself, lies at the foundation of the " Isthmus Method " 
of determining permeabilities under very high excitations 
introduced by Ewing and Low.* According to my some 
what extended experience the so called " maximum value of 
/" may be determined by the Isthmus Method to within 1 
or 2 per cent of the truth if the poles and the test-piece are 
of the proper shape and are properly connected, and if the 
jaws as well as the isthmus are fairly soft; but these conditions 
are not always easy of attainment, and if one assumes them to 
be satisfied without investigating each case by itself, one may 
be led into grievous error. The field about a hard steel 
isthmus between soft jaws is usually far from uniform, and for 
some specimens of hard steel which I have studied I have not 
yet succeeded in obtaining by the Isthmus Method trustworthy 
determinations of 7^ . In some instances the values of Pi in 
the air were manifestly smaller or larger than in the isthmus, 
and sometimes they were smaller for one excitation and larger, 
for the same isthmus, for another excitation. Nevertheless 
the method is, of course, a most valuable one. 

This paper describes a long series of determinations of the 
permeabilities of normal brands of compressed steel, Bessemer 
steel, and Norway iron in the form of half-inch rods, over a 
wide range of excitation, and it considers especially a method 
of measurement in a massive yoke in the interesting region 
from H = 400 to PL = 2500 which lies above the limits of 
most permeameter observations and below those of the Isthmus 
Method. The work was undertaken in order to determine the 
magnetic behavior, in the region just mentioned, of an impor- 
tant specimen of soft iron of which only a single short piece 
was available, and it was necessary to test the trustworthiness 
of the method to be used by applying it to some soft metals 
which could be obtained in large pieces and the permeabilities 
of which could be otherwise found, at least approximately. 

If the rod to be experimented upon can be kept cool artifi- 
cially, it is not very difficult to determine accurately in a very 

* Ewing, Magnetic Induction in Iron and other Metals ; Ewing and Low, 
Philosophical Trans., clxxx, 1889. 



276 Peii'ce — Permeabilities and Reluctivities for Steel. 

long solenoid, the permeability of a soft, homogeneous rod four 
or five hundred diameters long, for excitations up to H = 400. 
Here the value of I is probably between 95 per cent and 
97 per cent of the final value ( 1^ ), which can be found by the 
use of the Isthmus Method. If, then, any other method . of 
measuring permeability be used on the specimen between 
H — 300 and i/ = 2000 and if this method yields the proper 
values of I at both ends of the interval, it is comparatively easy 
to judge, from a graphical representation of all the observa- 
tions, whether the short interval corresponding to from 3 
to 5 per cent of 1^ is properly bridged. The principal 
difficulty with this procedure is that several isthmus specimens 
of a metal and. several testing coils must be used before one can 
be satisfied that the resulting value of 7^ is correct to within 
1 per cent ; for, as Ewing has shown, the separate results of a 
series of determinations of 7^ by the Isthmus Method may 
differ from the mean on either side by as much as 4 per cent. 
As a matter of fact, in all the materials I have tested the final 
value of /obtained by the Isthmus Method does not differ by so 
much as 1 per cent from the final value as obtained by the other 
method I have used. This second method, however, gives in 
any case a series of determinations of 1^ which do not as a 
rule range over so much as 1/3 per cent of the mean, while 
the Isthmus Method in my hands is much less satisfactory in 
this respect. 

The Bessemer and the compressed steel were procured in 
specially long pieces from which lengths of about 450 centi- 
meters were cut, and these, under the severest test which I 
could conveniently apply, seemed to be practically uniform 
throughout. The Norway iron, on the contrary, was not every- 
where of the same temper and could not be used satisfactorily 
until it had been carefully annealed by Mr. George W. Thomp- 
son, the mechanician of the Jefferson Laboratory, who has 
had much experience in work of this kind. Upon each of these 
rods at its center a test coil of fine wire was w T ound by Mr. 
John Coulson, who has made all the test coils and has helped 
me in every part of the work, and then the coil and its leads 
were carefully covered by pieces of rubber and rubber tape to 
make the whole waterproof. The rod thus prepared was 
placed inside a horizontal solenoid nearly 500 centimeters long, 
placed perpendicular to the meridian. This solenoid was 
uniformly wound with 20,904 turns of well insulated wire and 
a stream of tap water could be kept running through the bore 
around the rod to prevent any appreciable rise of temperature. 
The rod was demagnetized in situ by means of a long series 
of currents in the solenoid, alternating in direction and grad- 
ually decreasing in intensity, and then a curve of ascending 



Peirce — Permeabilities and Reluctivities for Steel. 277 

reversals was obtained in the usual manner. It is to be noticed 
that the demagnetizing process does not succeed unless the 
rod is practically homogeneous throughout its length. 

The ballistic galvanometer employed in this work has been 
already described* in a previous paper, and it is only necessary 
to say here that its period was so long that no detectable error 
was introduced into its indications by the fact that four or 
five seconds were necessary to make the magnetic changes 
corresponding to a reversal of current in the exciting coil of 
the yoke. For the largest currents a battery of about 120 large 
storage cells was used ; a battery of 30 or 40 cells furnished the 
weaker currents. 

When it is necessary to create inside an iron rod a fairly 
uniform magnetic field of intensity much greater than, say 
500, some kind of yokef is almost indispensable, and many 
kinds of yoke-permeameters are now used successfully in 
studying the magnetic properties of short rods at commercial 
excitations. For very high excitations, at which the air is 
nearly as permeable as the metal, the leakage becomes very 
troublesome and depends upon matters which cannot be easily 
controlled. I hoped, notwithstanding this fact, to be able to 
calibrate the yoke we had (figure 3) by means of standard 
pieces and thus make it available for studying short pieces of 
iron at excitations of about iZ=1000, but Mr. John Coulson 
and I worked for a long time on this problem without finding 
any satisfactory solution. We found it possible, however, to 
determine the length of a half-inch rod of iron which, mounted 
between the jaws of this particular yoke, would cause the mag- 
netic field in the air just without the iron near the middle of the 
specimen to be practically straight and parallel to the rod for 
a considerable distance. A piece about 15 centimeters long, 
accurately fitted for about 3*5 centimeters at each end into 
the taper holes in the jaws and leaving 80 millimeters of the 
rod free, satisfied the conditions, and this length might be 
slightly varied, but a much shorter rod violated the conditions 
in one direction (the determination of fi being too large) and 

* Peirce, Proc. American Academy Arts and Sciences, xliv, 1908. 

f H. E. J. G. DuBois, Phil. Mag. 1890, The Magnetic Circuit in Theory 
and Practice ; Shuddemagen,- Proceedings American Acad, of Arts and. 
Sciences, xliii, 1907 ; J. Hopkinson, Philosophical Transactions, clxxvi, 1885; 
Drysdale, Electrician, xxviii, 1901 ; Thornton, Electrician, xlix, 1902 ; 
Ewing, Electrician, xxxvii, 1896; Schmidt, Ann. der Physik, liv, 1895 ; 
Chattock, Electrician, xxxvii, 1896 ; Lamb and Walker, Electrician, xlvii, 
1901 ; Baily, Electrician, xlviii, 1901 ; Kopsel and Kobinson, Electrische 
Zeitschrift, xv, 1894; Kath, Electrische Zeitschrift, xix, 1898 ; Kennelly, 
Electrische Zeitschrift, xiv, 1893; Blondel, Comptes Eendus, cxxvii, 1898 ; 
Stoletow, Ann. der Physik, cxlvi, 1872; Eowland. Phil. Mag., xlvi, 1878 ; 
Ewing, Magnetic Induction, 1892 ; Behn-Eschenberg, Electrische Zeit- 
schrift, xiv, 1893 ; Kapp, Electrical Engineer, xxiii, 1894. 



278 jPeirce — Permeabilities and Reluctivities for Steel. 



a longer rod in the other. 



If, then, a testing coil (K) of very 
fine wire were wound in a single layer over a centimeter or 
two of the middle of a rod of the standard length, and a second, 
similar, coil (L) of radius about two millimeters greater than 
that of K, upon an extremely thin non-magnetic spool slipped 



Fig. 1. 



P 



B 



H 



Tm 



over K, it was to be expected that a knowledge of the whole 
amounts of the currents induced in Kand L when the exciting 
current of the yoke was reversed would determine, as in the 
Isthmus Method, corresponding values of R and B. This 
assumption was fully justified by experiment. The radii of 



Fig. 




the coils and the diameter of the wire were measured with the 
help of a Zeiss comparator (No. 3196). The coil L was 
wound upon a very dry piece of boxwood, and was carefully 
baked in shellac. Parafnne wax is inadmissible as an insulator 
on the wire because of its magnetic properties, and for the 
same reason a vulcanite spool cannot be used for L. 



Peirce— Permeabilities and Reluctivities for Steel. 279 
Fig. '6. 

j4l -* 1 




The moment of inertia of the suspended coil of the long 
period ballistic galvanometer used for the work was so great 
tha at Tow excitations the instrument was rather insensitive 
when used to measure the difference of the fluxes through 



280 Peirce — Permeabilities and Reluctivities for Steel. 

K and L, and it became necessary to make the scale distance 
nearly four meters. Under these circumstances it was prac- 
tically impossible with a very large reading telescope of the 
very best make (Clark's) to get a bright image of the scale suf- 
ficiently magnified to render ballistic determinations easy, so 
we had recourse to the simple device'" represented by 
figures 1 and 2, and this has given great satisfaction. In front 

Fig. 4. 




of the reasonably plane mirror (M) of the galvanometer, 
instead of the usual cover glass, was placed a convex spectacle 
lens (A) of about four meters focal length. At a distance in 
front of A equal to its focal length was a horizontal scale (S) 
mounted in the usual manner on a thin strip of wood at least 
twice as wide as the scale itself. Through the middle of this 
strip above the scale was bored a round hole (H) rather more 
than 20 millimeters in diameter, and just behind the scale a fine 
vertical wire or silk fiber (W) was stretched across the opening 
to serve as a cross hair. Behind H, and at a distance suited to 
its focal length and to the eye of the observer, was placed 
another spectacle lens (B) to serve as an eyepiece. This had a 
focal length of about 20 centimeters. A peephole (P) on the 
common axis of H, A, and B, so placed that B's aperture 
appeared wholly filled with a large clear uncolored image of 
the scale with the crosswire running vertically across it, com- 
pleted the arrangement. If different persons used the device 
the position of B had to be changed to suit the eyes of the 

* Peirce, Proc. American Academy Arts and Sciences, xlii, 1906. 



Peirce — Permeabilities and Reluctivities for Steel. 281 

observer, but the position of H was fixed. In setting up 
apparatus of this sort, one should remember that if the distance 
between B and P is not properly chosen, only a small round 
portion of the image of the scale will be visible, not nearly 
large enough to fill the aperture of B. 

The electromagnetic yoke used in the experiments described 
here is represented in figure 3. The three exciting coils are 
wound upon brass spools and these are mounted upon pieces 



Fig. 5. 

































s^V 




























^ 


^ 




























„ 


/'' 




























„ 


^y 




























Y^ 


^' 




















1 
1 










y^ 


' 




























/ 


'" 


























\ 


s" 


/ 
































































1 


J 

































Fig. 5. Curve showing the reluctivity of the soft Bessemer steel rod for 
values of H between 3 and 160. 



of soft steel shafting carefully turned to fit the holes in the 
heavy castings which complete the frame. The coils have 
together 2956 turns and the wire of which they are made is 
large enough to carry a current of 50 amperes for a few 
minutes at a time without undue heating. The whole yoke 
weighs about 300 kilograms. The apparatus is supplied with 
jaws of a number of different forms, but for the results 
recorded below conical jaws of the form shown in figure 4 
with taper holes to receive the tapered ends of the specimen 
to be tested were used. All the joints were made by Mr. 
Thompson and seemed to be mechanically perfect. Before 
each piece was tested the jaws were driven together with the 
test-piece between, and then the whole was clamped. 

For very low excitations the field about the test-piece did 
not seem to be quite uniform, but this difficulty disappeared 
when the exciting current became as strong as one ampere. 

The intensities of the exciting currents were measured with 
the help of a set of five Weston amperemeters of different 



282 Peirce — Permeabilities and Reluctivities for Steel. 

ranges, all but one of which could be shunted at pleasure out 
of the circuit. The currents passed through a rack of three 
rheostats specially made for the purpose by the Simplex 
Electric Company : these had a range of about 8000 ohms. 

The ballistic galvanometer was calibrated at short intervals 
by means of a coil (X) of 13*6 ohms resistance always in its 
circuit. This coil, X, consisted of 534 turns of fine insulated wire 
wound in a single layer upon a very accurately cylindrical 
wooden core which was hung within a uniformly wound ver- 
tical solenoid 176*2 centimeters long consisting of 5526 turns. 



Fig. 6. 

























\ 

pi 

1 






















1 






















1 

1 

4 






















1 

I 
1 
















J 




1o 


i 
\ 














>> 




/' 


T 




\ 








^> 




^ 


<' 








\ 


K ^ 


^> 


>"' 




r"*" 
















s> 


r 
















> 




<' 
















10 ; 


s 











































Fig. 6. The ordinates of the upper curve show the values of the reluc- 
tivity of the annealed Norway iron for values of H between and 10. The 
ordinates of the lower curve represent, on a scale one-tenth as large, the 
reluctivities of this iron for excitations between H = 10 and H= 100. 

The effective area of the turns of X was 22,720 square centi- 
meters and a current of one ampere sent through the solenoid 
caused a flux of 895,400 maxwells through X. 

The remarkable solenoid, nearly five meters long, in which 
the permeabilities were determined for values of II below 410, 
was constructed by Mr. Thompson and his assistants in a lathe 
the bed of which had been temporarily lengthened. The core 
was a thick-walled solid-drawn brass tube of about one 



Peirce — Permeabilities and Reluctivities for Steel. 283 

inch inside diameter made by the American Tube Works. 
The solenoid was mounted on a long oak truss which kept it 
from buckling. 

Every specimen was tested in the solenoid several times 
over to make sure that the demagnetizing process was so 
effective that at low excitations, where differences are often apt 
to appear, the same results were obtained at every trial. Until 
the Xorway iron had been specially annealed, it was impossible 



Fig. 7. 



F 


* 


























U.I 
























* 


C 
























f 
























s 
s 


s 
s 

r 
























s 


/ 






















J 


s 

s 

r 
























s 


/ 
























> 

* 


s 
























J 


s 

f 
























j 

* 

s 


























G 
























1 























20 


50 







Fig. 7. This curve shows the reluctivity of the annealed Norway iron for 
values of H between 100 and 2200. 



so to demagnetize the rod that for low currents the induction 
should always be exactly the same. As Shuddemagen* has 
shown, the "end correction" for a round straight rod is a 
function of the absolute dimensions of the rod and not merely 
of the ratio of the length to the diameter, but the correction 
in the case of a half -inch rod of the length' used in the work 
here described, is very small. 

No other explanations seem necessary to make the following- 



tables intelligible. 

* Shuddemagen, Proc. Amer. Acad. 
DuBois, The Magnetic Circuit, § 31. 



Arts and Sciences, xliii, 190" 



28-Jr Peirce — Permeabilities and Reluctivities for Steel. 



Table I. — Bessemer Half -Inch Bod in Long Solenoid. 


H 


B 




H 


B 
















0-5 


115 






10 


5 10650 


1-0 


280 






11 


10890 


1-5 


465 






15 


12400 


2-0 


730 






20 


13360 


2-5 


1180 






30 


14480 


3-0 


1935 






40 


15200 


3-5 


2800 






50 


15720 


4-0 


3760 






60 


16120 


4-5 


4830 






70 


16460 


5-0 


5700 






80 


16750 


5-5 


6410 






90 


17000 


6*0 


7060 






100 


17220 


6-5 


7650 






120 


17640 


7-0 


8130 






160 


18240 


7'5 


8600 






200 


18800 


8-0 


9040 






240 


19200 


8-5 


9420 






270 


19620 


9-0 


9760 






300 


19800 


9-5 


10070 






350 


20240 


10-0 


10390 






400 


20660 


Table II.— 


Half-Inch Bessemer Bo 


i in Massive Yoke. 




(Free 


length about 80 mm ) 


H 




B 




I 


400 




20660 




1613 


500 




21250 




1652 


600 




21660 




1676 


707 




21900 




1686 


800 




22020 




1688 


1000 




22230 




1689 


1208 




22470 




1692 


1500 




22770 




1692 


2090 




23550 




1692 


2590 




23 


860 




1693 



Numerous observations were made for values of H lying 
between 350 and 1600 and these were used to draw a curve 
from which some of the mimbers given above were taken. 
Above 1 600, I had only two observations and the results of 
these for H— 2090 and iZ=2590 appear at the end of the 
table. The observations for H—107 and iZ"— 1208 were the 
only ones in these neighborhoods and are given as they stand. 
The value of I for iZ=2590 happened to come out 1693, but 
a study of the observations shows that this number is really 
uncertain by about two units in the last place, for B is not 
determined in the fourth significant place. 



Peirce — Permeabilities and Reluctivities for Steel. 285 



An isthmus, cut from this rod gave for a long series of 
observations for values of iTup to 16300, the same final value 
for /, within a fraction of 1 per cent, as the results tabulated 
above. 

In view of the importance of Frohlich's Law for magnetic 
metals at commercial excitations, figure 5, which shows the 



Fig. 































































.j 


r- 


■N 


M 












j 




< 






1 

1 








> 


> 


r 










l 
1 

\ 




J 


y 
>" 


<■ 














1 

V 


i 


^ 








































1 
























O 10 H 

Fig. 8. This curve shows the reluctivity of the compressed steel between 
£r=3jind jB"=100. 

reluctivity of this specimen of soft Bessemer steel for values 
of ^H up to 160, is of some interest. 

Table III. — Half-Inch Rod of Norway Iron in Long Solenoid. 



H 


B 


H 


B 












1-0 


1960 


24 


16180 


2-0 


6950 


28 


16320 


3-0 


9780 


30 


16400 


4-0 


11380 


40 


16650 


5-0 


12400 


50 


16920 


6'0 


13080 


60 


17180 


7-0 


13640 


70 


17400 


8-0 


14140 


80 


17600 


10- 


14800 


100 


17940 


12- 


15200 


120 


18220 


14- 


15460 


160 


18800 


16- 


15620 


200 


19100 


18- 


15780 


240 


19840 


20- 


15960 


280 


20200 



At low excitations, the determinations of B for this rod 
were a little uncertain because of the difficulty of demagnetiz- 



286 Peirce — Permeabilities and Reluctivities for Steel. 

ing the specimen completely. Investigation showed that 
there were very slight differences of temper at different parts 
of the rod, and it seemed best to have the iron thoroughly 
annealed. This process increased the permeability for almost 
all excitations, very materially, as a comparison of Tables III 
and IV will show. 

Table IV. — Annealed Half-Inch Norway Iron Hod in Long 

Solenoid. 



H 


B 






H 


B 
















0-2 


190 






9 


5 14800 


0'5 


395 






10 


14940 


0-8 


1120 






11 


15100 


1-0 


2160 






12 


15360 


1-5 


4600 






14 


15540 


2-0 


6600 






16 


15700 


2-5 


8240 






18 


15900 


3-0 


9480 






20 


16040 


3'5 


10460 






25 


16320 


4-0 


11280 






30 


16520 


4-5 


11980 






. 35 


16740 


5-0 


12560 






40 


16920 


5-5 


13000 






45 


17100 


6-0 


13400 






50 


17220 


6'5 


13700 






60 


17450 


7-0 


13900 






70- 


17630 


7*5 


14100 






80 


17820 


8-0 


14300 






90- 


18020 


8-5 


14490 






100- 


18210 


9-0 


14660 






105- 


18300 


Table V. — Annealed Norway Iron Ho 


d in 3Iassive Y< 




(Free 


length about 80 mm 


) 


H 




B 




I 


295 




21120 




1657 


350 




21550 




1687 


400 




21850 




1707 


450 




22100 




1723 


500 




22220 




1729 


600 




22480 




1741 


700 




22630 




1745 


800 




22770 




1748 


900 




22880 




1749 


1000 




23000 




1750 


1500 




23500 




1751 


1800 




23810 




1751 


2000 




24010 




1751 


2350 




243 


so 




1751 



Peirce — Permeabilities and Reluctivities for Steel. 287 



Figures 6 and 7 show the reluctivity of this iron at differ- 
ent excitations. 

Two other half-inch specimens of very pure Norway iron 
turned from two-inch bars obtained from different sources 
gave as maximum values of the magnetization vector (I) 1732 
and 1738. These were in the condition in which they were 
bought and had not been specially annealed. 



Table Yl.—Halj 


'-Inch Mod of Compressed Shafting in Ion 






Solenoid 








H 


B 




H 




B 












13- 




10400 


0-5 


70 






14- 




10690 


1-0 


155 






15- 




10970 


1-5 


290 






20* 




12300 


2'0 


490 






25' 




13260 


2-5 


800 






30- 




13950 


3*0 


1280 






40* 




15150 


3'5 


2010 






50' 




15850 


4-0 


2880 






• 60* 




16480 


4'5 


3210 






70- 




16950 


5-0 


4450 






80* 




17320 


5 5 


5120 






90- 




17640 


6-0 


5730 






100 




17900 


6-5 


6270 






120 




18200 


7-0 


6780 






140 




18560 


7-5 


7200 






160 




18800 


8-0 


7610 






180 




19080 


8*5 


8020 






200 




19320 


9-0 


8380 






240 




19700 


9-5 


8690 






280 




20000 


io- 


9010 






320 




20360 


ii- 


9540 






360 




20600 


12- 


10100 






400 




20820 


Table VII.— Short Piece 


of Half 


■Inch 


Compressed Shafting 




Magnetized in Massive 


Yoke. 






(Free 


length about 80 n,m ) 




H 




B 






I 


300 




20220 






1585 


370 




20630 






1612 


400 




20820 






1625 


500 




21200 






1647 


600 




21500 






1663 


700 




21670 






1669 


800 




21810 






1672 


1000 




22080 






1678 


1200 




22340 






1682 


1600 




22800 






1687 


1854 




23 


080 






1690 



288 Peirce — Permeabilities and Reluctivities for Steel. 

Figure 8 shows the reluctivity of this steel for excitations 
below #=100. 

Table VIII gives the results of an interesting study made 
by Mr. Coulson of the magnetic properties of a piece of 
Kidd's polished tool steel, for excitations between i7=250 
and #=2500. 







Table VIII. 






H 


B 


/" 


E 


I 


300 


17470 


58-2 


0-017 


1364 


350 


17750 


50-7 


0-020 


1385 


400 


17980 


44-9 


0-022 


1399 


500 


18370 


36-7 


0-027 


1422 


600 


18710 


31-2 


0-032 


1441 


800 


19240 


24-0 


0-042 


1468 


1000 


19680 


19-7 


0-051 


1487 


1200 


20050 


16-7 


0-060 


1500 


1400 


20440 


14-6 


0-068 


1515 


1600 


20790 


13-0 


0-077 


1527 


2000 


21400 


10*1 


0-094 


1544 


2500 


22010 


8-8 


0-114 


1549 



The line obtained by plotting the reluctivity (R) against H 
has only a very gentle curvature between these limits for this 
annealed tool steel. 

I wish to express my obligations to the Trustees of the 
Bache Fund of the National Academy of Sciences for the 
loan of some of the apparatus used in making the measure- 
ments recorded in this paper. 



The Jefferson Physical Laboratory, 

Harvard University, Cambridge, 



Mass. 



W. J. Miller — Ice Movement and Erosion. 



289 



Art. XXI. — Ice Movement and Erosion along the South- 
western Adirondacks y* by William J. Miller. 



Introduction. 

While engaged in field work along the southwestern border 
of the Adirondacks during the past three summers, the writer 
has had occasion to study some of the glacial features so well 

Fig. 1. 




Fig. 1. Sketch map of the southwestern border of the Adirondacks. 
Arrows indicate direction of ice movement. 

shown in that region. The area here described extends from 
Lowville. Lewis Co., southwestwardly past Boonville, Oneida 
Co., to Dolgeville, Herkimer Co., a distance of about 60 miles. 

* Published by permission of the New York State Geologist. 

Am. Jour. Sci.— Fourth Series, Vol. XXVII, No. 160.— April, 1909. 
20 



290 W. J. Miller — Ice Movement and Erosion, 

The width varies greatly but averages about 15 miles. The 
following topographic maps of the TJ. S. Geological Survey 
cover most of the region : Port Leyden (aud northward and 
eastward), Pemsen, Little Falls, together with portions of the 
Wilmurt, Boonville, and Utica quadrangles. (See sketch map.) 

The chief topographic feature of the northern portion of the 
area is the Black river valley through which Black river flows 
in a northwesterly direction. This valley is from 10 to 15 miles 
wide and shows a maximum depth of about 1300 feet. West 
Canada creek, the principal tributary of the upper Mohawk 
river, traverses the southern part of the area. This southern 
part is very hilly and shows a general slope towards the Mohawk 
river. The watershed separating Black river and West Canada 
creek represents one of the principal divisions of drainage in this 
part of the state. Leaving the Mohaw T k river at Utica (eleva- 
tion 400 ft.), and passing northward along the R. W. & O. R. R., 
there is a gradual ascent to the divide (elevation 1280 ft.) 
south of Alder creek. From this point northward there is a 
gradual descent along the railroad to near Lowville, where 
Black river shows an elevation of 740 ft. The common range 
of altitudes along both the eastern and western sides of the 
district is between 1200 ft. and 2000 ft. 

The Paleozoic-Precambrian boundary line passes lengthwise 
through the region. The Precambrian rocks comprise chiefly the 
highly metamorphosed Grenville sediments, the post-Grenville 
syenite-gneiss, and large areas which are more or less intimate 
mixtures of the Grenville and syenite. The Paleozoics overlap 
on the Precambrians and exhibit an excellent section showing 
the Pamelia sandy limestones, Lowville limestones, Black 
river shales and limestones, Trenton limestones, Utica shales, 
Lorraine shales and sandstones and the Oswego sandstone. 
The type localities of the Lowville, Black river and Trenton 
are found here. 

The full significance of the glaciology of this territory will 
not be known until the contiguous areas have been carefully 
studied. The discussion here presented is local in its charac- 
ter, the purpose being to record observations and to offer some 
conclusions which may aid in the solution of the broader prob- 
lem. So far as the writer is aware, nothing has been published 
regarding the northern portion of the district, while Cham- 
berlin,* Brigham,f and Gushing;}: have each referred to the 
southern portion. Brigham is now engaged in the study of 
the Mohawk valley region. § 

* Third Annual Eep. U. S. G. S., 1881-2, pp. 360-365. 

fBull. Geol. Soc. Am., vol. ix, pp. 183-202, 1898. 

JN. Y. State Museum, Bull. 77, pp. 73-81, 1905. 

§ Since the above was written several papers bearing on the glacial history 
of Central and Northern New York have been read by H. L. Fairchild before 
the 1908 meeting of the Geological Society of America. 



W. J. Miller — Ice Movement and Erosion. 291 

Direction of Ice Movement. 

Chamberlin, in the report above referred to, makes the ten- 
tative statement" that massive ice currents having their ulterior 
channels in the Cham/plain valley, on the one hand, and the 
Saint Lawrence on the other, swept around the Adirondacks 
and entered the Mohawk valley at either extremity, while a 
feebler current, at the height of glaciation, probably passed 
over the Adirondacks and gave to the whole a southerly trend." 
Observations by later investigators have tended to bear out this 
view, and the evidences from the southwestern Adirondacks 
herewith presented have an important bearing upon the propo- 
sition. 

The direction of flow is best shown by the glacial striae 
which have been observed at a number of different places 
through the district. The striae are best preserved upon the 
hard Precambrian rocks, but these are mostly drift-covered 
except along the chief stream courses. The limestones are 
next most favorable, while upon the shales none have been 
found. Striations are present only upon those surfaces from 
which the drift has been recently removed, because even the 
hardest rocks exposed during postglacial times have been 
weathered enough to cause an obliteration of the glacial marks. 

Striae pointing from south 25° to 40° east have been located 
as follows : On Trenton limestone one mile south of Martins- 
burg and also one-third of a mile east of Martinsburg (S. 25° E.) ; 
on Precambrian near the mouth of Roaring brook; on Pre- 
cambrian one and one-half miles northeast-north of Olenfield, 
and also one-third of a mile southwest and three-fourths of a 
mile southeast-south of the same village ; on Precambrian one 
mile northeast of Denley station, one-third of a mile north- 
east of Hawkinsville, on Big "Woodhull creek two miles north- 
east of Forestport, and one mile north of Salisbury centre 
(Cushing). Other striae, bearing nearly south, occur on Pre- 
cambrian one mile east of Port Leyden and still others bearing 
]ST. 80° E. on the Trenton limestone two and a half miles north 
of Middleville (Cushing). Beside these the observations of 
Chamberlin, in the Mohawk Valley near Little Falls, are 
quoted : " On the western slope of the ridge that stands athwart 
the valley of the Mohawk at Little Falls are two sets of striae. 
The course of the main series is S. 50° E. and of the minor 
cross-set S. 45° W. North of Little Falls, two-thirds of the 
way up the slope of the valley the course of the striae is S. 60° 
E., the movement being apparently from the west. A little 
farther up the slope, on an eastward inclined rock surface, the 
main set runs S. 50°-55° E. crossed by feebler and later ones 
S. 70° W. Four miles north of Little Falls, a varying group 
runs S. 18°-28° E. and about six miles north S. 37° E." 



292 W. J. Miller — Ice Movement and Erosion. 

The strise described above and plotted on the map (fig. 1) 
indicate that in the northern part of the region here discussed 
the southeasterly movement changed to the more nearly easterly 
movement farther south, and this is just what would be 
expected according to the statement of Chamberlin. It should 
be noted that the Black river and upper Mohawk valleys, 
which are the chief topographic features of the district, had 
much to do with determining the direction of the flow of the 
ice. Both of these valleys existed in preglacial times and the 
close parallelism between the directions of the striae and the 
directions of the valleys shows the influence of the latter in 
determining the ice movement. 

There is other evidence, derived from the distribution of the 
drift, to show the general southeasterly ice movement in the 
region. Thus, within the Remsen quadrangle north of Hinck- 
ley, fragments of Utica shale and Lorraine sandstone may be 
found at least five or six miles east of the nearest parent ledges. 
This strongly suggests a southeasterly current of ice. Also 
the presence of marginal morainic materials and delta deposits 
extending many miles along the southwestern Adirondacks 
show that a border ice mass existed there during the time of 
melting. Along the northwestern border of the Adirondacks 
the ice undoubtedly moved southwestwardly. The writer has 
seen magnificent displays of glacial grooves and strise bearing 
southwestwardly, especially near Clayton. Along the eastern 
border of the Adirondacks, the general southerly movement 
of the ice has been well established, as has also the westerly 
movement up the Mohawk valley towards Little Falls. Thus 
the statement of Chamberlin, regarding ice flow around the 
Adirondacks, harmonizes almost perfectly with observed strise. 

But the question still arises, what was the direction of the 
current during the height of glaciation ? We have abundant 
evidence to prove that this main current was a southwesterly 
one. The Long Lake quadrangle is located in the midst of 
the Adirondacks and upon the geological map of that area 
Prof. Cushing* has recorded a number of striae, all of which 
point toward the southwest. Over the region south of the 
Adirondacks and the Mohawk valley the numerous observa- 
tions of both Brighamf and Chamberlin;}; show that the ice 
moved in a general southwesterly direction. Along the south- 
western Adirondacks the writer has seen a number of good 
examples of roches moutonnees, all of which bear southwest- 
wardly. Among these one forms a narrow ridge about two 
miles long which crosses Moose river above Fowlersville (Port 

*N. Y. State Museum, Bull. 115, p. 495, 1906. 

\ This Journal, vol. xlix, p. 216, 1895. % Op. cit., p. 365. 



. W. J. Miller — Ice Movement and Erosion. 293 

Leyden sheet), while other good ones occur at an elevation of 
1700 feet, two miles northwest of Reed's Mill (Pemsen sheet). 
Another strong evidence favoring the southwesterly current is 
the distribution of glacial bowlders over the region southwest 
of the Adirondacks. Most of the common Adirondack rock 
types are strewn over the region and they gradually diminish 
in number as the distance from the mountains, becomes greater. 
This subject has been discussed in a paper by Brigham.* 

Thus, bearing in mind all the facts, the writer is strongly of 
the opinion that when the ice in its movement (during the 
last invasion) struck the Adirondacks, it was divided into two 
currents flowing around the mountains and meeting in the 
Mohawk valley ; that during the time of maximum glaciation 
there was a strong general current, but that the border cur- 
rents continued as under currents (more or less checked in 
velocity); and that, after the disappearance of the ice sheet 
from the central Adirondacks, border currents were main- 
tained. 

Ice Erosion - . 

It seems to be generally agreed upon by those who are 
familiar with the region that the topography of northern New 
York was not profoundly affected by ice erosion. Certainly 
no valleys of any consequence have been formed by ice action. 
Prof. Fairchild gave the testimony of Gushing, Smyth, and 
Gilbert to this effect in a paper read before the Geological 
Society in 1905. f The writer is in agreement with this general 
view, but he could not consent to the idea that the ice did 
practically no work of erosion in northern New York. 

1. Erosion of the Precambrian Pocks. 

As the ice moved across the region discussed in this paper, 
the preglacial rock surface was more or less scratched, polished, 
and eroded. In the case of the Precambrian rocks, it is doubt- 
ful if the ice did any very deep cutting. The work of erosion 
involved mostly the removal of joint blocks of the decayed and 
weathered rock materials near the surface. The evidence is 
conclusive that the weathered materials were rather thoroughly 
scraped off the Precambrians, as shown by the remarkable 
freshness of the rocks wherever exposed and by the smoothed 
and rounded character of the outcrops. Wherever the streams 
have cut through the mantle of drift and into the underlying 
Precambrians the surfaces of the latter are very hard and fresh. 
The highly jointed character of these rocks no doubt greatly 
aided the ice in its work of erosion. In this connection men- 

* This Journal, vol. xlix, pp. 213-228, 1895. 

f Bull. Geol. Soc. Amer., vol. xvi, pp. 50-5, 1905. 



294 W. J. Miller — Ice Movement and Erosion. 

tion should be made of the great number of large, more or less 
jointed, to somewhat rounded and fresh erratics of Precambrian 
rock material strewn over much of the region, especially the 
central portion of the Remsen quadrangle. It is very com- 
mon to find erratics ranging from five to twenty feet in diame- 
ter, one at least measuring seventeen feet high and twenty- 
seven feet across. The larger ones are mostly of the hard, 
homogeneous syenite or granite. Probably the greatest amount 
of ice erosion of the Precambrians occurred along Black river 
between Lyons Falls and Lowville ; but this matter will be 
referred to below. 

2. Erosion of the Sedimentaries. 

Turning our attention to the sedimentary rocks, we find that 
ice erosion was much more effective upon them. In fact, the 
writer believes that in the Black river valley we have one of 
the best examples of ice erosion in northern New York. One 
factor favoring the ice work here was the comparative softness 
and the highly jointed character of the rocks, while another 
factor was their ^position with reference to the ice current (see 
below). The writer has found no sign of considerable ice ero- 
sion in the southern part of the region here described. 

Fig. 2. 



Fig. 2. Section across the Port Leyden quadrangle two and one-half 
miles north of Lyons Falls. Vertical scale 8*8 times the horizontal scale. 

The accompanying figure shows the profile and the geologic 
structure across the Black river valley two and one-half miles 
north of Lyons Falls. One of the striking features is the ter- 
raced character of the sedimentaries, particularly from Port 
Leyden northward. (See topographic map.) Along the river 
course there is a slight notch in the Precambrians and just west 
of this, on the northern part of the Port Leyden sheet, there is 



W. J. Miller— Ice Movement and Erosion. 295 

a steep slope rising three hundred feet above the Precambrians. 
The formations outcropping on this slope are shown in, the sec- 
tion. Resting upon the Precambrians are several feet of weak 
sandstones which are followed by the sandy limestones of the 
Pamelia; then come the hard Lowville and Black river lime- 
stones, followed by weak lower Trenton shales and limestones ; 
while the summit is capped by the hard, crystalline Trenton 
limestones. The streams passing over this slope are character- 
ized by gorges with waterfalls and rapids. From the summit 
of this slope and extending for several miles westward is a well 
defined terrace developed upon the limestone. 

Rising from the western side of the above-named terrace 
there is a second slope higher and much steeper than the first. 
The rise is commonly about 450 feet within a third of a mile. 
The soft Utica shales outcrop at the base of this slope and 
they are followed by the Lorraine shales with an upward 
increasing sandstone content. The summit of this terrace, 
known as Tug Hill, is more irregular and stream-dissected 
than the limestone terrace below. A]l streams flowing across 
the steep slope of this terrace have high gradients and have cut 
deep narrow gorges locally called " gulfs." A third terrace, 
much less well-defined, is formed by the Oswego sandstone 
capping. The much higher gradients of the eastward flowing 
streams from Tug Hill and the much steeper slope and greater 
height of the eastern front of Tug Hill^ as compared to other 
parts of the district, all argue for a recent and considerable 
cutting back (westward) of the Paleozoics here. 

At first these terraces, in their present form, were thought 
to have been due entirely to water action, but an examination 
of the region shows that some other explanation must be 
sought. The steep fronts of the terraces are certainly young 
topographic features, which precludes the possibility of their 
having been formed during the long preglacial period of 
erosion in this ancient region. On the other hand, Black river 
has done very little work of erosion, between Lyon Falls and 
Lowville, in postglacial times, as proved by the fact that the 
stream has not yet cut its way through the alluvium and 
reworked drift filling the valley bottom, and also because 
glacial striae and kames near the river level have not been dis- 
turbed. Thus also the slight trench cut into the Precambrians 
along here could not have been postglacial in origin. 

There is still the possibility that glacial waters might have 
developed the terraces, but the writer has looked in vain for 
evidence of any such vigorous water action, especially along 
the higher part of the limestone terrace where records would 
surely be left. Even if a large stream had flowed along the 
ice edge and under the steep front of Tug Hill, its gradient 



296 W. J. Miller — Ice Movement and Erosion. 

would have been too low to be 'compatible with much cutting 
power. No doubt there was some movement of water along 
the edge of the waning Black river ice lobe, but the only cur- 
rent of any importance was a northerly one between the 
eastern edge of the limestone terrace and the ice margin. 
The limestones here are somewhat waterworn, but this 
stream was about 200 feet below the top of the terrace and 
thus clearly could not have done the work of erosion over 
the whole terrace. Also the presence of the glacial strise high 
up on the terrace shows that no great amount of water erosion 
could have taken place there since the ice retreat. 

The Black river valley certainly existed, in its broader out- 
line, in preglacial times. The Paleozoic-Precambrian bound- 
ary line had for a long time been gradually moving westward 
by wearing away of the Paleozoics. It seems certain that the 
lowermost Paleozoic layers must have extended farther east- 
ward, by overlap on the Precambrians, immediately preceding 
the glacial period. This means that Black river was some dis- 
tance farther eastward and that the western tributaries, from 
Tug Hill, entered it with lower gradients. As above shown, 
the lowest sedimentary layers could not have been cut. back in 
pre- or post-glacial times nor were they cut back by glacial waters. 
Evidently they were cut back by the ice to develop the steep 
slope now shown. This allowed Black river to shift westward to 
its present position. Thus the slight trench of the Precambrians 
here shown could not have been preglacial. As already stated, 
it is clearly not post-glacial and apparently it was formed by ice 
cutting. The concave character of this inner portion of the 
valley is well shown in the figure and strongly suggests ice 
work. 

Also we should consider the fact that we are here dealing 
with unaltered sedimentaries with slightly upturned edges 
resting upon a rather smooth surface of igneous and metamor- 
phic rocks, and that the lowest sediments are w T eak sandstones 
and sandy limestones, which greatly favored the stripping off 
power of the ice. Robert Bell* has noted the same thing in 
Canada, and he says : " When unaltered strata lie at low angles 
upon a nucleus of crystalline rocks, there is a marked differ- 
ence in the effects produced by the action of the passing ice- 
sheet according as the latter moved from the overlapping 
strata onto the solid nucleus or off the latter against the 
upturned edges of the stratified rocks, . , . in the latter (case), 
great erosion has always taken place and valleys and basins are 
formed whose width depends upon the angle of dip and the 
softness of the strata which have been scooped out. The strata 

*Bull. Geol. Soc. Amer., vol. i, p. 296, 1890. 



W. J. Miller — Ice Movement and Erosion. 297 

are presented in the most favorable attitude for abrasion. The 
wearing down would go on until the resisting rock front had 
attained a height and weight sufficient to counterbalance those 
of the glacier. " In the Black river valley the ice moved from 
the crystallines against the slightly upturned edges of the sed- 
iments. 

In much the same way the soft shales were stripped off the 
surface of the hard limestones to form the broad terrace 
and the steep front of Tug Hill. Such a stripping off of the 
shales occurred, but to a less extent, over the southern part of 
the Port Leyden quadrangle (except at Locust Grove; ; over 
the northwestern part of the Boonville quadrangle; and over 
the western part of the Pemsen quadrangle. This power of 
erosion diminished southward. The maximum thickness of 
shale thus removed was probably several hundred feet, but 
not over a wide area. The total amount of shale removed 
was not nearly as much as may at first sight be supposed. 
Then, too, the shales were soft and highly jointed, even to a 
considerable depth, as may now be seen in the Whetstone gulf 
section. 

It may be fairly asked, what became of the materials thus 
removed ? The very resistant Precambrians ought to be pres- 
ent in considerable force as erratics somewhere in the region 
and this is the case especially in the townships of Boonville and 
Pemsen, where a vast number of such erratics may be seen 
strewn over the country. Shale and limestone is also present 
in great abundance in the till and other drift of the Pemsen 
quadrangle. However, much of the shale must have been 
ground up and carried away by glacial waters. 

Two other factors which greatly aided the work of the ice 
in' the Tug Hill region must not be overlooked. One of these 
is the fact that the ice moved up hill as it advanced southward 
along the valley and so had its cutting power increased. On 
reaching the divide between Black river and West Canada creek 
the cutting power was lessened and till and other drift mate- 
rials were deposited in great quantities as the ice moved down 
hill toward the Mohawk river. Another factor which the 
writer regards as important in this connection is the angle at 
which the ice current entered the Black river valley in its 
sweep around the Adirondacks. The greatest amount of 
erosion was along the eastern side of Tug Hill, and it was just 
here where the ice current must have struck with greatest force 
as it was crowded into the valley. In harmony with this idea 
is the fact that the glacial striae near Martinsburg bear more 
towards the south than does the steep front of Tug Hill. 

It appears to the writer that, even under these favorable 
circumstances, the ice did not show itself to be such a great 



298 W. J. Miller — Ice Movement and Erosion. 

erosive agent as claimed by some geologists. The conclusion 
seems to be that the only considerable ice erosion was along 
the eastern side of Tug Hill and the valley eastward, where the 
ice caused channel-straightening accompanied by some deepen- 
ing. The very soft shales were considerably cut away ; the 
harder limestones were less affected ; while the very hard Pre- 
cambrians yielded least. Gilbert* in speaking of ice erosion in 
western New York, near Lock port, says : " The district exhibits 
a marked contrast between the extent of erosion from a broad 
mass of limestone on the one hand and a broad mass of shale on 
the other, the ratio being, roughly, as 1 to 10 or 1 to 20 ." 
Thus in the Black river valley we also have a good example 
of differential ice erosion. 

Hamilton College, Clinton, New York. 
January, 1909. 

*Bull. Geol. Soc. Amer., vol. x, p. 180, 1899. 



Edgar — Estimation of Vanadic and Arsenic Acids. 299 



Art. XXII. — The Estimation of Vanadic and Arsenic Acids 
and of Vanadic and Antimonic Acids, in the Presence of 
One Another ; by Graham Edgar. 

[Contributions from the Kent Chemical Laboratory of Yale Univ. — cxcvi.] 

Arsenic and Vanadium. 

The almost constant association of arsenic and vanadium in 
the mineral sources of both elements presents the frequent 
problem of their separation and estimation and has led to the 
development of numerous analytical processes to accomplish 
the desired end. Carnot* precipitates all arsenic acid by boil- 
ing with a salt of strontium in weakly ammoniacal solution in 
the presence of ammonium salts. Gribbsf precipitates the 
arsenic as trisulphide from a solution previously reduced with 
sulphur dioxide, the vanadium being determined in the filtrate 
by titration with potassium permanganate. Sehmitz-Du- 
montj uses hydrogen sulphide under pressure to effect the 
separation. Frieclheim and Michaelis, § after reducing with 
sulphur dioxide the solution containing arsenic and vanadic 
acids, separate the former by repeated distillation with methyl 
alcohol and hydrochloric acid. Field and Smith [ volatilize 
the arsenic by heating the dry sulphides of arsenic and vana- 
dium in a current of hydrochloric acid gas at a temperature 
between 100° and 250° Centigrade. Friedheim, Decker and 
Diem^f recommend the volatilization of the arsenic by distilla- 
tion with potassium iodide and hydrochoric acid, a current of 
hydrogen being kept up through the apparatus during the 
process. 

In the present investigation any separation of arsenic and 
vanadium is avoided by the use of a process of differential 
reduction to determine these elements in the presence of one 
another. 

If a solution containing arsenic and vanadic acids is boiled 
with tartaric or oxalic acids, the vanadic acid is reduced to 
tetroxide and may be reoxidized in alkaline solution by iodine*"* 
according to the equation 

V 2 4 + I 2 + H 2 = V 2 5 + 2HI 

If a solution containing arsenic and vanadic acids is reduced 
with sulphur dioxide under proper conditions, the arsenic acid 

*Compt. Bend., civ, 1803. fChem. Jour., vii, 230. 

{Inaug. Diss., Berlin, 1891. 

§ Berichte tier Deutsch. Chem. Gesellsch., xxviii, 1414. 
|| Journ. Amer. Chem. Soc, xviii, 1051. 
""Zeitschr. Anal. Chem., xliv, 648. 

** Browning: Zeitschr. anorg. Chem., vii, 158; Browning and Goodman, 
this Journal, ii, 355. 



300 Edgar —Estimation of Vanadic and Arsenic Acids. 

is reduced to arsenious acid and the vanadic acid to tetroxide, 
so that after boiling off the excess of reagent the reoxidation 
by iodine in alkaline solution should proceed according to the 
equation 

As 2 3 + V 2 4 + 3l 2 + 3H 2 0=As 2 B + V 2 5 + 6Hi 

and the iodine thus used should correspond to the sum of 
the two oxides. If then aliquot parts of the same solution 
are treated in the manner described above, it is evident that 
the titration of the solution reduced by tartaric acid should 
determine the vanadium present and that the titration of the 
solution reduced by sulphur dioxide should determine the sum 
of the arsenic and vanadium, the difference in the number of 
cubic centimeters of iodine used in the two cases being the 
amount required to oxidize the arsenic. 

In Table (I) are given the results of experiments performed 
upon solutions containing arsenic and vanadic acids in varying 
proportions. The method of treatment in detail w T as as fol- 
lows : The solutions were divided into two portions aud one of 
these was boiled with one to two grams of tartaric or oxalic acid 
until the blue color of the vanadium tetroxide indicated complete 
reduction. The solution was then cooled, nearly neutralized 
with potassium bicarbonate and an excess of standard iodine 
solution was added. Neutralization was then completed, an 
excess of bicarbonate added, and the solution allowed to stand 
for from fifteen minutes to one half hour. The excess of 
iodine was then removed with standard arsenious acid and the 
solution titrated to color after the addition of starch. The 
results of this titration are given under (I), Table (I). 

The second portion of the solution was placed in a small 
pressure flask and slightly acidified with sulphuric acid. 
Twenty-five cubic centimeters of a strong solution of sulphur- 
ous acid was then added and the flask was closed and heated for 
one hour on the steam bath.* After cooling, the flask was 
opened and the solution transferred to an Erlenmeyer flask 
and boiled to remove the excess of sulphur dioxide, a current 
of carbon dioxide being passed into the liquid to facilitate the 
removal of the last traces. The solution was then cooled, 
nearly neutralized with potassium bicarbonate and an excess 
of iodine added as before. After completing the neutralization, 
adding an excess of bicarbonate and allowing to stand for one 
half hour, the excess of iodine was determined by arsenious 
acid as before. The results are given in (II), Table (I). 

As before stated, the results in (I) determine the vanadium, 
and these figures, subtracted from (II), determine the arsenic. 

*McCay: Amer. Chem. Jour., vii, 273. 



Edgar — Estimation of Vanadic and Arsenic Acids. 301 
Table (I). 



Taken 
V 2 5 


Found 
V 2 5 


Error 

v 2 o 5 


Taken 

As 2 5 


Found 
As 2 5 


Error 
As 2 5 


(I) 

N/10 

Iodine 


(II) 

N/10 
Iodine 


grm. 


grm. 


grm. 


grm. 


grm. 


grm. 


cm 3 . 


cm 3 . 


0-1183 


0-1181 


— 0-0002 


0-0960 


0-0961 


+ 0-0001 


12-95 


29-65 


0-1183 


0-1183 


±0-0000 


0-0960 


0-0962 


+ 0-0002 


1297 


2970 


0-1183 


0-1182 


— o-oooi 


0-0960 


0-0962 


+ 0-0002 


12-96 


29-70 


00591 


0-0593 


+ 0-0002 


0-0480 


0-0480 


±0-0000 


6-50 


14-85 


0-0591 


0-0594 


+ 0-0003 


0-0480 


0-0482 


+ 0-0002 


6-52 


14-90 


0*0591 


0-0589 


— 0-0002 


0-0480 


0-0483 


+ 0-0003 


6-45 


14-86 


0-1774 


0-1779 


+ 0-0005 


0*1440 


0-1438 


— 0-0002 


19-50 


44-50 


0-1774 


0-1774 


±0-0000 


0-1440 


0*1440 


±0-0000 


19-45 


44-50 


0-1774 


0-1776 


+ 0-0002 


0-1440 


0-1442 


+ 0-0002 


19-47 


44-45 


0-2366 


0-2371 


+ 0*0005 


0-0480 


0-0478 


— 0-0002 


26-00 


34*31 


0-2366 


0-2366 


±0-0000 


0-0480 


0-0480 


±0-0000 


25-95 


34-30 


00591 


0-0593 


+ 0-0002 


0-1440 


0-1439 


—o-oooi 


6*50 


38-90 



Antimony and Vanadium. 

Since antimonic acid is reduced by sulphur dioxide on 
heating in a pressure flask* there seemed no reason to suppose 
that antimony and vanadium should not be determined in the 
presence of each other by a process similar to the one used for 
arsenic and vanadium. Accordingly a series of experiments 
was made in which solutions containing vanadic acid and anti- 
monic acid were treated in a manner exactly similar in detail 
to the preceding. The results are given in Table (II). 



Taken 


Found 


s Error 


Table 
Taken 


(II). 
Found 


Error 


(I) 
N/lOx 
0-9375 


(II) 
N/lOx 
0-9375 


V 2 5 


v 2 o 5 


V.O5 


Sb 2 5 


Sb 2 5 


Sb 2 5 


Iodine 


Iodine 


grm. 


grm. 


grm. 


grm. 


grm. 


grm. 


cm 3 . 


cm 3 . 


0-1183 


0-1185 


+ 0-0002 


0-0757 


0-0759 


+ 0-0002 


13-85 


23-69 


0*1183 


0-1186 


+ 0-0003 


0-0757 


0-0764 


+ 0-0007 


13-87 


24-05 


0-1183 


0-1183 


±0-0000 


0-0757 


0-07G0 


+ 0-0003 


13-83 


23-95 


0-1774 


0-1777 


+ 0-0003 


0-1261 


0-1258 


— 0-0003 


20-80 


37-55 


0-1774 


0-1777 


+ 0-0003 


0-1261 


0-1258 


— 0-0003 


20*80 


37-55 


0-1774 


0-1773 


—o-oooi 


0-1261 


0-1260 


—o-oooi 


20-75 


37-50 


0-2366 


0-2376 


+ 0*0010 


0-1261 


0-1257 


— 0-0004 


27-80 


44-52 


0-2366 


0-2369 


+ 0-0003 


0-1261 


0-1263 


— 0-0002 


27-70 


44-50 


0-2366 


0-2369 


+ 0-0003 


0-1261 


0-1260 


—o-oooi 


27-70 


44*45 



Summary. 

In the preceding paper it has been shown that vanadic and 
arsenic acids, or vanadic and antimonic acids, may be readily 
determined in the presence of one another by a method based 
upon the differential reducing action of tartaric or oxalic acid 
and sulphur dioxide, the reoxidation being in each case effected 
with iodine in alkaline solution under the conditions described 
above. 

*VonKnorre : Zeitschr. angewandte Chem. 1888, 155. 



302 Goooh and Bosworth — Iodometric Estimation of Silver. 



Art. XXIII. — A Method for the Iodometric Estimation of 
Silver Based upon the Use of Potassium Chromate -as a 
Precipitant y by F. A. G-ooch and Rowland S. Bosworth. 

[Contributions froni the Kent Chemical Laboratory of Yale Univ. — cxcvii.] 

It lias been shown in a previous paper that with proper pre- 
cautions silver may be precipitated and accurately estimated as 
silver chromate. It was found that the addition of a sufficient 
excess of potassium chromate to a solution of silver nitrate, 
even in the presence of small amounts of nitric acid, brings 
about a complete precipitation of the silver as silver chromate, 
and that the precipitate thus obtained may be transferred to 
the asbestos filter by means of a dilute solution of potassium 
chromate and washed with small amounts of water without 
appreciable loss of silver chromate. Upon the basis of such 
exact precipitation of silver chromate by potassium chromate 
it should be possible to establish a method for the iodometric 
estimation of silver either by the estimation of the chromic 
acid ion of the precipitated and washed silver chromate or by 
the determination of the chromic acid ion of the potassium 
chromate which remains after the precipitation of the silver 
salt by a known amount of standard potassium chromate. 

In studying the latter procedure a known amount of stand- 
ard potassium chromate in excess of the amount needed to 
precipitate the silver was added to the solution of silver nitrate. 
The precipitate was dissolved in ammonia and reprecipi- 
tation brought about by boiling to a volume of 10-15 cm \ The 
second, crystalline precipitate was filtered upon asbestos and 
washed with the least possible amount of water applied in small 







Table I. 






Silver taken. 










Volume 










Error in 


of solu- 




K 2 Cr0 4 


Na 2 S 2 3 


Silver 


terms of 


tion 


Weight 


used 


used 


found 


silver 


cm 3 . 


grm. 


grm. 


cm 3 . 


grm. 


grm. 


20 


0-1261 


0-3039 


26-54 


0-1262 


+ 0-0001 


25 


0'1576 


0-3039 


22-61 


0-1575 


—o-oooi 


15 


0-0946 


0-3293 


30-51 


0-0946 


o-oooo 


15 


0-0946 


0-3293 


30-60 


0-0940 


— 0-0006 


15 


0-0946 


0-3293 


30-57 


0-0941 


— 0-0005 


15 


0-0946 


03293 


30-55 


0945 


— 0-0003 


19 98 


0-1260 


0-3293 


32-20 


0-1255 


— 0-0005 


20 


0-1261 


0-3293 


32-09 


0-1263 


+ 0'0002 


20 


0-1261 


0-3293 


32-10 


0-1263 


+ 0-0002 


25 


0-1576 


0-3293 


27-30 


0-1576 


o-oooo 



Gooch and Bosworth — lodometric Estimation of Silver. 303 

portions successively. The filtrate was treated with potassium 
iodide and acidified with sulphuric acid. The iodine set free 
was titrated with sodium thiosulphate. The difference be- 
tween the silver value of the iodine thus found and that of the 
potassium dichromate used was taken as the measure of the 
silver present. In Table I are given the details of experiments 
performed in accordance with this procedure. 

In Table II are given the details of similar experiments in 
which the precipitation was effected in the presence of sodium 
nitrate. 









Table II. 








Silver taken. 












Volume 




K 2 Cr< 






Error in 


of solu- 
tion 






> , NflKOo 


Na 2 S 2 3 
used 


Silver 
found 


terms of 
silver 


Weight 


Volume 


*\ J-ICUJNW3 

Weight present 


cm 3 . 


grm. 


cm 3 . 


grm. grm. 


cm 3 . 


grm. 


grm. 


10 


o-noi* 


37 


02436* 1 


20-47* 


0-1107 


— 0-0006 


10 


o-noi 


37 


0-2436 1 


20-53 


0-1103 


+ 0-0002 


10 


o-iioi 


25 


0-1647 1 


9-07 


0-1097 


— 0-0004 


10 


o-noi 


27 


0-1778 1 


11-02 


0-1096 


— 0-0005 


15 


0-0862f 


+ 30 


0-1974* 2 


17-14* 


0*0859 


— 0-0003 


15 


0-0862 


30 


0-1974 1 


17-05 


0-0865 


+ 00003 


15 


0-0862 


30 


0-1974 1 


17-14 


0-0859 


— 0-0003 


25 


0-1437 


50 


0-3294 10 


28-53 


0-1435 


— 0-0002 



* Approximately N/10. 



f Approximately N/20. 



Since, as has been shown in the paper previously referred 
to, relatively large amounts of potassium chromate are neces- 
sary to bring about complete precipitation of the silver chro- 
mate in the presence of nitric acid, the above procedure 
seemed less adapted to the determination of silver in a solution 
containing that acid than the method whereby the precipitated 
and washed silver chromate is determined. The following is 
an account of the results obtained in the study of this latter 
procedure. To the silver solution containing free nitric acid, 
potassium chromate was added in excess of the amount necessary 
to take up the nitric acid with formation of potassium dichrom- 
ate. The precipitate was dissolved in ammonia and reprecipi- 
tation effected by boiling to a volume of 10-15 cm3 * The second, 
crystalline precipitate was transferred to an asbestos filter by 
means of a dilute solution of potassium chromate, washed with 
the least possible amount of water applied in small portions 
successively, and dissolved in a few cm 3 of a strong solution of 
potassium iodide. The solution in potassium iodide was diluted 
and acidified with sulphuric acid. The iodine set free was 
titrated with sodium thiosulphate and taken as the measure of 



304 Gooch and Bosworth — lodometric Estimation of Silver. 

the silver present. In Table III are given the details of exper- 
iments made in this manner. 









Table IN. 






Silvei 


• taken. 












Volume 












Error in 


of solu- 




HN0 3 


K 2 Cr0 4 


Na a S a O» 


Silver 


terms of 


tion 


"Weight 


present 


weight 


used 


found 


silver 


cm 3 . 


grm. 


grm. 


grm. 


cm 3 . 


grm. 


grm. 


25 


0-1348 


0-063 


0-60 


18*35 


0-1342 


— 0-0006 


20 


0-1078 


0-063 


0-65 


14-67 


0-1073 


— 0-0005 


15 


0-0808 


0*063 


0*65 


11-06 


0-0809 


+ 0-0001 


15 


0-0808 


0-063 


0-65 


10-96 


0-0802 


-0-0006 


20 


0-1078 


0-063 


0-65 


14-67 


0-1073 


— 0-0005 


30 


0-1618 


0-063 


0-75 


22-02 


0-1610 


— 0-0008 


20 


0-1078 


0063 


0-65 


14-71 


0-1075 


— 0-0003 


25 


0-1348 


0-063 


0-65 


18-41 


0-1347 


— 0-0001 


25 


0-1348 


0-063 


0-65 


18-41 


0-1347 


— 0-0001 


20 


0-1078 


0-063 


0-65 


14-74 


0-1078 


o-oooo 



The results obtained justify the conclusion that silver can be 
accurately estimated by precipitating as silver chr ornate with 
the use of a sufficient excess of potassium chromate, dissolving 
the precipitate formed in ammonia, reprecipitating by boiling 
to low volume, and determining iodometrically either the chro- 
mate ion in combination with the silver or the chromate ion 
of the potassium chromate remaining after precipitation of 
the silver with a known amount of standard potassium 
chromate. 



Ileadden — Brown Artesian Waters of Costilla Co., Colo. 305 



Art. XXI Y. — The Brown Artesian Waters of Costilla 
County, Colo., their Relations to Certain Deposits of 
Natron or Soda, and what they teach / by ¥m. P. Headden. 

The San Luis Valley, in the southern part of this state, 
presents several interesting questions, among them its drainage 
and its artesian wells. The fact that there is an artesian basin 
in this valley was discovered by accident in 1887. I have no 
official figures at my command in regard to the number of such 
wells in the valley, but I have heard it estimated at more than 
3000. They are put down with such ease that it is difficult to 
obtain accurate records, either in regard to their number or the 
strata passed through by the borings. The strata consist of 
alternate layers of clay and sand which scarcely attains to the 
size of fine gravel. The depth at which flows are met with 
varies exceedingly. South and west of La Jara, water is 
struck at very shallow depths ; a four-inch, cased well on the 
place of Mr. Ormond is 77 feet deep, the water rises a few 
inches above the casing and is very cold, 44° F. Even shal- 
lower wells than this are recorded, which is no cause for sur- 
prise as it is very probable that some of the springs in the 
eastern and southern parts of the valley are due to the artesian 
pressure in the valley and not to surface waters finding an 
exit in the usual manner. The ground and. surface waters of 
the valley are, so far as my knowledge goes, of bad quality, 
whereas these springs and the artesian wells yield excellent 
waters. Even wells of 1000 feet depth yield waters carrying 
less than 16 grains to the imperial gallon and 45 per cent of 
this mineral matter is silicic acid. The springs and wells 
agree in the amount of total solids carried in solution, ranging 
from 5*8 to 20*7 grains to the gallon, and are characterized by 
a high percentage of silicic acid, namely, from 26 to 46 per 
cent of the total solids. 

The percentage of flowing wells obtained is very large ; fail- 
ures are very rare, but an occasional one is met with. Mr. W. 
K. Hapney informs me that he put down a well, a few miles 
west of Alamosa, to the depth of 1004 feet, without obtain- 
ing a flow — at least the well was a failure. This is certainly 
remarkable, for at Alamosa they have two of the strongest 
flows in the valley. It is difficult to obtain full enough data 
to justify one in venturing a suggestion in explanation of such 
a failure. There are no reasons for supposing that the strata 
are not continuous throughout this section, nor have we any 
reason, except perhaps this failure itself, for assuming that a 
fold may exist at this place. 

Am. Jour. Sci.— Fourth Series, Vol. XXVII, No. 160.— April, 1909. 
21 



306 Headden — Brown Artesian Waters of Costilla Co., Colo. 

The supply of water is furnished by the streams descending 
from the surrounding mountains. The artesian basin has a 
length of about 75 miles and a width from 40 to 45 miles. 
The Rio Grande del Norte enters the valley at a point about 
the middle of its western side and then flows easterly and 
southerly out of the valley by way of its canyon. All of the 
streams which enter the valley north of the Rio Grande are 
wholly lost in the sands of the valley, their waters do not at 
any time join those of the Rio Grande by overground courses 
and it is a question whether they do so, at least within the 
limits of the valley, by underground circulation. So complete 
is the disappearance of the waters of these northern streams 
that it is only very rarely that the united waters of the 
Saguache river and the San Luis creek have been known to 
reach so far south as the San Luis lake, 14 miles north and 7 
miles east of the town of Alamosa. The Rio Grande loses 
from 75 to 100 second feet of water on its entrance to the val- 
ley. The streams to the south also lose large quantities of 
w T ater as they enter the valley or daring their course through 
it. These sources are probably quite sufficient to supply all 
the artesian waters of this basin. 

In the eastern part of the northern half of this basin we 
find a water wholly different from any of the springs or wells 
previously mentioned. The water heretofore described has 
been a colorless, tasteless water, carrying but small amounts of 
mineral matter in solution and the chief constituent of these 
mineral matters is silica. In this area, which begins about eight 
miles north of Alamosa, and extends almost to the town of 
Moffat, a distance of 25 to 30 miles from north to south and 
extending from 5 or 6 miles west of the town of Mosca 
almost to the San Luis lake on the east, a distance of about 
14 miles east and west, the water has a decidedly " red " or 
brown color, generally smells of hydrogen sulphide, is often 
accompanied by marsh gas, has an alkaline reaction and is rich 
in sodium carbonate. The wells sunk in this area show these 
characteristics throughout. The flows struck at about 200 feet 
are slightly brown, taste of hydrogen sulphide, and contain 
about 22*0 grains of solid matter per imperial gallon, which is 
wholly sodium carbonate. The intensity of \\iq color deepens 
with the depth of the wells till at about 500 feet it attains its 
maximum ; from this point the color diminishes till at about 880 
feet it has become somewhat fainter but still decidedly colored. 
The mineral matter, in this case sodium carbonate, increases with 
the depth ; at 200 feet it is 22*0 grains, at 500 feet 76'0 grains, 
and at 800 feet 103 4- grains in each imperial gallon. No 
well sunk in this section has passed through this zone of red 
or brown water and obtained white or colorless water beneath 



Headden — Brown Artesian Waters of Costilla Co., Colo. 307 

it. On the southeastern limit of this area there seems to be 
one instance in which an area of colorless water is overlapped 
by this area of brown water, but this is exceptional and while 
easily explicable if true, needs confirmation. 

The brown water area is surrounded by an area of colorless 
water and the limits, at some points, are fairly well defined. 
I observed two wells near the southern limit of the brown 
water ; one 500 feet deep yielding a brown water carrying 
104'0 grains total solids per gallon ; less than two miles to the 
south of it another, 880 feet deep ; here the water was scarcely 
colored at all and carried only 38 grains per gallon ; this well 
emitted a considerable quantity of combustible gas, but was 
free from hydrogen sulphide, and the water was nearly taste- 
less. The wells which yield gas most freely seem to be at the 
outer edge of the brown water area and have considerable 
depth. I do not know of any well close to the San Luis lake 
on the west side, but the nearest one that I now recall yields a 
brown water and a fair quantity of gas ; its depth is, according 
to my information, upwards of 900 feet. At the northeast 
corner of the lake and close to it there is a 500-foot? well, 
which yields a white water, tasting slightly of hydrogen sul- 
phide but otherwise of good quality. A short distance south 
of this, and directly east of the lake, is a well 300 feet deep, 
which furnished a colorless water, but this well is no longer 
flowing. 

It is as good as impossible to obtain the logs of these wells, 
for they have been sunk through clay and sand with such ease 
and rapidity that it would be somewhat difficult for anyone to 
make an accurate record, which may be more fully appreciated 
when it is considered that wells as much as 400 feet deep have 
been begun and finished in a day. Some fragmentary informa- 
tion is furnished by the debris washed out of the drill holes 
such as is shown by the presence of fish vertebrse and, within 
the brown water area, fragments of wood. The presence of 
the wood suggests an explanation for the brown color of the 
water, i. e., that there is enough vegetable matter buried here 
to give rise to a sufficient quantity of humus to color it. 

One of the questions arising in connection with this valley 
is, what becomes of the water flowing into it? This valley 
receives the water from a large drainage area, but the visible 
discharge of water from it amounts to only 540,000 acre feet 
annually or about 75 second feet, which is less than the loss of 
the Rio Grande as it enters the valley. I have not been able 
to obtain even an approximate idea of the total amount of the 
water entering the valley, but it is evidently very much in 
excess of the visible outflow. 



308 Headden — Brown Artesian Waters of Costilla Co., Colo. 

These conditions, stated imperfectly and in general terms, 
raise the question of a general underground flow out of the 
valley. 

We have, in the case of these brown waters, an artesian 
basin of approximately 420 square miles. The water, finding 
its way into this area, must be brought into the valley by the 
streams descending from the surrounding mountains which 
carry only snow and rain water. In the other portions of the 
valley the artesian waters are not only pure, but have the 
characteristics of the water which we find in mountain streams; 
in this area we find them characterized by a decided color, the 
presence of hydrogen sulphide, accompanied in some cases by 
marked quantities of combustible gases, probably marsh gas, 
and containing sodium carbonate. Further, these waters are not. 
confined to one or more horizons within this area, but all flows, 
from the most shallow to the deepest, bear the same character- 
istics, and the area itself has quite sharply defined limits. If 
there be any general movement of these waters, it must be to 
the southward, for the valley is completely enclosed by moun- 
tains except on the south. The pressure shown by these wells 
is in no case great. The Denver and Rio Grande R. R. had a 
well put down at Alamosa to a depth of about 1,000 feet. 
The flow is excellent, but the pressure is not sufficient to force 
the water into the tank for supplying the engines, so it is prob- 
ably less than 25 feet. Such a pressure with a head of 360 
feet, this being the approximate difference of level between 
iUamosa and the western portion of the valley, is not incom- 
patible with an outflow to the southward, perhaps a hundred 
or more miles away. 

As the characteristics of these brown waters may aid in the 
solution of these general questions, I have paid some attention 
to their study. 

If these waters were to pass through strata very different 
from those of the section in which they occur, the brown color 
would in all probability be removed, but not so with the sodium 
carbonate, for even clayey soils have but a limited power to 
retain sodium carbonate, as is shown by the fact that it passes 
quite freely into drain waters from soils which contain but 
little of this salt. 

The sanitary analysis of the brown water presented some 
difficulty due to the organic matter present, but less than the 
water from the San Luis lake. In the former case the 
organic matter is probably humus, in the latter it is entirely 
different, as it is without color and at most imparts a slight 
milkiness to the water. The question of a possible connection 
of the San Luis lake with the reservoir of artesian water is a 
proximate one ; indeed it is said that the water has been seen 



Heaclden — Broion Artesian Waters of Costilla Co., Colo. 309 

to well up in the middle of the lake, accompanied by a 
gurgling sound. It is easy to conceive of such a connection, 
especially if the springs in the eastern and southern parts of 
the valley owe their origin to the artesian pressure in the basin 
rather than to individual local causes. This view, on the other 
hand, has but little plausibility when applied to the Head 
lake, a small adjacent lake, for this lake becomes almost 
entirely dry at times. These lakes fill up in the spring time, 
occasionally receiving water from the Saguache river and the 
San Luis creek, though this is of very rare occurrence, and 
have no visible outlet. I do not know the rate of evaporation 
from a quiet surface at this place, but it must be high, 
because the sunshine is almost continuous and the temperature 
in the summer season is high, and high winds are of frequent 
occurrence. 

I have observed differences of level in the waters of the San 
Luis lake, varying to possibly 10 feet. The cause of this is not 
entirely clear ; it seems too great a difference to attribute to 
surface evaporation. If the water of this lake is supplied from 
an underground source, it probably owes its origin to more 
immediate surface drainage, rather than to artesian pressure. 
This inference is drawn from the character of the mineral 
matter held in solution rather than from knowledge of strati- 
graphical relations. The analysis of this lake water gives the 
following characteristics : First, it is comparatively rich in 
mineral matter, 54*2 grains of fixed residue in each imperial 
gallon. Second, it is exceptionally rich in potash salts, 
potassium carbonate constituting one-fifth of the total, while the 
corresponding soda salt constitutes one-fourth. Third, the 
determination of saline and albuminoidal ammonia present 
exceptional difficulties. 

While the amount of the total solids, 54'2 grains per 
imperial gallon, seems high, a little consideration of the condi- 
tions removes this impression. This lake has no visible outlet; 
it is not known how long it has already existed here, and the 
accumulation of salts, due to the concentration of pure moun- 
tain or snow water, might account for all of the salts that we 
find. Prof. L. G. Carpenter informs me that the evaporation 
from the surface of this lake may reasonably be estimated at 
sixty inches per annum, at which rate the water contained in 
the lake at any one time would be wholly evaporated in about 
three- years. This is allowing an average depth of about 15 
feet, a much greater depth than the lake possesses. Such con- 
siderations lead to the conclusion that the mineral matter held 
in solution is not sufficiently large to preclude their having 
been wholly derived from surface waters. . The presence of so 
large a percentage of potash salts is more consistent with a 



310 Headden— Brown Artesian Waters of Costilla Co., Colo. 

surface origin than a deep-seated one, for while the waters of 
our mountain streams are not so rich, even relatively, in 
potash salts as this lake water, they approach it much more 
nearly than any others known to me. In cases where the 
watersheds are composed of granites, gneisses and schists, in 
which the predominant feldspar is orthoclase, we find the waters 
characterized by a low mineral content consisting of much 
silicic acid and a relatively large amount of potash. These 
two characteristics are present in this water, the former, how- 
ever, not in a very marked degree, but the latter strongly so. 

The question whether the silicic acid may have separated 
from this water as a difficultly soluble silicate is one on which 
the deportment of similar water under pressure, for instance 
when used in steam boilers, may throw a little light. In sev- 
eral cases which have come under my observation I have found 
the incrustation formed to consist very largely of silica. In 
one case in which the boiler had been in service for four years 
and had been fed with artesian water coming from two differ- 
ent flows bnt of essentially the same character, the incrustation 
formed on the boiler tubes was one quarter of an inch in 
thickness and consisted of silicic acid and lime, 76 per cent of 
the former and 24 per cent of the latter, including a small 
amount of alkalies. Some such separation may have removed 
a part of the silicic acid from the lake water, in which there is 
at the present time only a small amount of lime, one grain per 
gallon. 

The ammonia determinations in the sanitary analysis of this 
water presented peculiar and greater difficulties than I had 
ever met with in a water. The water is colorless and not 
strongly alkaline. There is no suggestion of the presence of 
humus : -the slowness and persistency with which the ammonia 
distilled over was the only suggestion of its presence. It has 
been suggested that when successive, equal portions of the dis- 
tillate show about one half as much ammonia as the preceding 
one, the probable presence of humus is indicated. According 
to this criterion the first three portions indicated the presence 
of humus, but each of the next six portions contained equal 
amounts of ammonia and no end of the reaction was obtained. 
In the distillation of the albuminoidal ammonia distilled water 
(ammonia free) was added, and 14 portions of 50 cc each were 
distilled over without obtaining an end to the reaction. The 
organic matter indicated by the slow and persistent evolution 
of ammonia may have come from a variety of sources, for the 
lake abounds in vegetation and, as I am credibly informed, in 
other aquatic life, a small fish attaining a length of Ave or six 
inches being especially abundant. As there is no outlet to this 
lake, it is evident that the water must become heavily charged 



Headden — Brown Artesian Waters of Costilla Co., Colo. 311 

with organic matter. These points in the composition and 
deportment of this water distinguish it from the spring and 
well waters. 

The well waters to the east of the lake are, so far as I know, 
white waters and of fairly good quality, but the wells to the 
west of it but not close to it yield brown waters ; those of the 
lower flows are so rich in sodium carbonate that they are no 
longer potable waters and but poorly adapted for boiler use 
owing to the fact that they foam badly. I have previously 
stated that these brown waters are characteristic of an area of 
approximately 420 square miles, and that all of the flows 
within this area have the same characteristics, namely, they are 
all more or less colored and all carry sodium carbonate. The 
flow met with at a depth of about 200 feet, in the town of 
Mosca, is used for domestic purposes. This water is slightly 
colored, tastes of hydrogen sulphide, and carries 22*4 grains of 
sodium carbonate in each imperial gallon. There are no sul- 
phates and only traces of chlorides present in this water. This 
flow is not strong, and shows a diminution as the number of 
wells tapping it is increased. The water from the town well, 
which was originally something over 800 feet deep, but in 
which the casing was subsequently pulled to the 500 or 600 foot 
point, is strongly colored, has a decided alkaline reaction, and 
contains 72*6 grains of fixed residue per gallon, of which over 
90 per cent is sodium carbonate. This is probably the most 
strongly colored water with which I have met. The mill well 
in the same town is from 780 to 800 feet deep, cased to the 
bottom ; the water from this well differs from the preceding 
only in containing more sodium carbonate or total solids per 
gallon, 102*8 grains, and in not being nearly so deeply colored. 
The sanitary analysis of these colored brown waters presented 
difficulties of the same nature, but much less serious than those 
presented by the San Luis lake water. They were so serious 
in both cases that I resorted to defecation by means of milk of 
lime, to which they were readily amenable. 

These brown waters are wholly different in their properties 
from the lake water, and also from other artesian water flowing 
from wells in comparatively close proximity ; for instance, the 
im socla well," a brown water, carried 103*0 grains per gallon ; 
the "gas well," supposed to be on the southern limit of the 
brown water area, carried 37*6 grains ; while the white water 
wells, a few miles south or east, carry only 6 grains. These 
brown waters also differ from the spring waters of the eastern 
part of the valley, which as a rule correspond closely in their 
character to the artesian waters of the basin. As an illustra- 
tion of this statement we have the Washington Springs, which 
carry 5*958 grains per gallon, and the nearest artesian well, of 
whose water I have an analysis, carries 6*698 grains per gallon. 



312 Ileadden — Brown Artesian Waters of Costilla Co., Colo. 

The composition- of these total solids is remarkably similar, 
especially in their high content of silicic acid. The residue 
from the well water contained 28*25 per cent, and that from 
the spring water 29'25 per cent of silicic acid. This similarity 
between the spring and well waters is not always so striking as 
in this case, but is, as a rule, very evident. The brown water 
area also has its springs, which are located immediately east of 
the San Luis lake, but the pressure is not sufficient to produce 
freely discharging springs, but instead small lakes of intensely 
alkaline water from which is deposited, both in winter and 
summer, considerable quantities of sodium carbonate, natron 
associated with a little trona. 

I have given the characteristic differences between the lake 
water and the brown artesian waters and shown that they are 
distinct waters. The lake water is colorless, and contains nota- 
ble quantities of silicic acid and almost as much potash as soda 
salts. The brown waters are distinguished from the other 
artesian waters of the basin by the presence of humus, the 
practical absence of silicic acid, and the presence of large quan- 
tities of sodium carbonate, from 23 to 103 grains to the imperial 
gallon. Further, these brown waters, or such as are rich in 
sodium carbonate, are confined to a quite sharply limited area, 
while the colorless waters are general throughout the rest of 
the artesian area. 

The colorless artesian waters of the basin carry only a very 
moderate amount of dissolved substances; even the water from 
the deepest wells contains less than 16 grains of total solids to 
the gallon, 46 per cent which is silicic acid. These facts lead 
us to conclude that these brown waters are the source of the 
natron found in these soda lakes, because neither the ordinary 
artesian waters of the valley nor the lake water nor the spring 
waters are either rich enough in soda or pure enough to, in 
any easily explainable way, give rise to such deposits of natron, 
while the brown waters might easily do so. 

The facts already adduced seem sufficient to justify the 
assumption that the natron found in these depressions owes its 
origin to the brown waters, but there are also other facts which 
support this view. I have submitted the water of the mill 
well, also those of the town well at Mosca, and of the mill well 
at Hooper, to examination with the purpose of comparing 
them with the mother liquor, collected in the so called Soda 
lake. Carefully made analyses of the residues from the three 
wells show that they are identical, so that I am justified in 
using a larger quantity of residue obtained from the mill well 
at Hooper for the detection of iodine, bromine, lithia, titanic and 
boric acids and in considering it as representative of the brown 
waters in general. The substances named were found in both 
the mother liquor from the natron and in the brown waters, 



Headden — Brown Artesian Waters of Costilla Co., Colo. 313 

but it was necessary to use as much as 20 grams of the residue to 
establish the presence of lithia and bromine. The iodine, titanic 
and boric acids, on the other hand, were easily detected. 

We have thus added to the fact that the brown waters are 
the only ones in the valley which, to our knowledge, carry 
sodium carbonate as the principal salt in solution, this further 
consideration, i. e., that the mother liquor from the crystals of 
natron resemble the brown waters in that it contains small 
quantities of iodine, lithia, titanic and boric acids. The potas- 
sium salts in the mother liquor of the lake are no more abundant 
than might be expected due to their concentration in this 
liquor as the soda salts crystallize out ; but there may be a 
question as to whether the iodine, etc., show an increase corre- 
sponding with the degree of concentration which has taken place. 

Apropos to the extent of these soda deposits. I may state 
that there are, to my knowledge, four depressions in which 
crystallized sodium carbonate, natron, is found. Three of these 
are quite small but the fourth is larger. When I last saw this 
there was probably a third of an acre entirely covered with a 
layer of natron, but it has at times been considerably larger 
than this. Mr. W. H. Falke, the present owner, tells me that 
he has measured the crystalline mass when it had a maximum 
thickness of forty-eight inches. There would have been no 
trouble in finding spots where the mass of crystals would have 
measured thirty inches in thickness when I last visited the place. 

There is a question which will certainly suggest itself to 
everyone in any way familiar with our western conditions, i. e., 
whether the surface waters of the section may have contributed 
to the formation of these deposits. It is well known that 
water percolating through our surface soils gives rise to 
efflorescences and incrustations on evaporation. I have never 
seen deposits owing their origin to this source, comparable to 
this in extent or thickness ; still it remains to show that these 
salts — this carbonate of soda did not come from the surface 
soil. I have analyzed the surface and also the ground waters 
from a territory which I believed would be most likely to give 
me results favorable to the assumption that such might be the 
source of the carbonate. The surface waters contain principally 
sulphates and chlorides with some carbonates ; the ground waters 
contain the same mixture, the sodium chloride being rather 
more abundant. The efflorescences occurring in this section 
are essentially sodium sulphate with subordinate quantities of 
calcium and magnesium sulphates. The aqueous extract of the 
soil contains large quantities of sulphates with only a little 
carbonate. We have then in the surface waters, the ground 
waters, the water-soluble portion of the soil, and in the efflor- 
escent salts gathering on the soil an entirely different mixture 
of salts and one which, were it to flow through or over the soil 



oli Headden — Broivn Artesian Waters of Costilla Co., Colo. 

into a basin and evaporate, would not give the pure sodic car- 
bonate found in this deposit. 

The natron was found in August, 1907, forming a mass of 
crystals live inches thick covered by an inch or more of a yel- 
lowish mother liquor. The upper part of the crystalline mass 
showed a thin layer of bright needles or thin prismatic crystals : 
these did not effloresce, yielded water in the closed tube and 
consisted of sodium carbonate. It was not feasible to select 
material pure enough for analysis. I took this layer as con- 
sisting of the mineral trona. 

The mass of crystals had effloresced so badly by the time I 
was able to analyze it, that I dried the material thoroughly 
before analyzing it. A water determination, however, was 
made on some of the best material that I could, under the 
circumstances, select and yielded 59*161 per cent of water. 

Analysis of thoroughly dried Analysis of the residue from 

salt-mass from Soda Lake. the mother liquor.* 

Si0 2 0-128 Si0 2 0-405 

S0 3 ..: 0-184 S0 3 2-235 

CI ._- 0-138 CI 2-409 

C0 3 41-060 C0 3 36-016 

Na,0 57-875 Na 2 46*991 

K 2 G__ 0-824 K 2 12-347 



100-209 100-403 

= C1... -030 = C1 ._ -543 



100-179 99 860 

*This residue contained traces of iodine and lithia, also of titanic and 
boric acids. 

The conditions obtaining in this artesian basin which 
permit of the confinement of these waters to a somewhat 
sharply limited area are certainly not clear. There are no 
reasons for assuming the existence of a separate basin in 
this section though its waters are very distinct in their proper- 
ties. There is no evident source of the sodium carbonate, for 
granting, as we must, that vegetation may have at some time 
been abundant in this section, which is indicated by the fossil 
wood, the humus and the marsh gas, it does not suffice to 
explain the presence of the sodium carbonate. We might 
assume that beds of carbonate exist somewhere in this terri- 
tory, but we have no proof that this is a fact. No observations 
indicating such deposits have been made in putting down the 
numerous wells of this section. The small deposits mentioned 
in this article are the only known ones and these are 
confined to the surface, or within a few feet of it. I have 
understood that borings have been made with the idea of 
finding such deposits, but without success. 



Headden — Brown Artesian Waters of Costilla Co., Colo. 315 

Analyses of Residues from Artesian Wells. 

Mosca. Town well Mosca. Mill well Hooper Mill well 
500-600 feet deep.* 780 feet deep. 750 feet deep. 

Carbon 0-365 

Si0 2 _. ,. 5'537 

S0 3 0-342 

CI 0-425 

P„0 5 0-247 

C0 2 37-603 

Na 9 0_.... 53-077 

K 2 G 0-982 

CaO 0-413 

MgO 0-153 

Fe 2 3 0-086 ) 

A1 2 3 0-121} 

Mn 3 4 0-086 

Li, I, Br ... . I , Li, etc. plus 

TiO„.BO .. h eactltiace Ore.m 



99-437 
= C1 0-096 



4-821 








4-160 


0-046 








0-063 


0*284 








0-470 


0-148 








0-168 


37-104 








37-365 


53-665 








54-045 


1-148 








1-433 


0-651 








0-282 


0-391 








0-219 


0-031 








0039 


0054 








0048 


LS 


Li 


, etc. 


pi. 


IS 


>r [1-721] 


Oi 


■g. matter [1.814 


100-064 








100-106 


0064 








0-106 



99-341 100-000 100-000 

* This residue was heated till the organic matter was charred but not 
wholly destroyed. The deficit of 0'6597 in the analysis is probably due to 
organic matter. 

If this water is characteristic of a subordinate basin inde- 
pendent of the rest of the valley, as I at one time thought, it 
is peculiar that we have no area surrounding it barren of 
artesian water. Further, that they have not succeeded in 
getting below the water carrying sodium carbonate in solution. 
There is no indication, of which I have been able to learn, that 
they have approached the downward limit of the sodium carbon- 
ate. The deepest well in this section, whose waters I have 
analyzed, was 880 feet deep and this water was the richest of 
all I have examined in sodium carbonate. On the other hand, 
if it forms a part of the general basin and there are no changes 
in the inclination of the strata, it is hard to see how the lateral 
limits of the sodium carbonate can be so sharply defined. There 
can certainly be no general lateral movement of the waters 
through the valley in which this water participates. Such a 
movement is possible, but it must take place at a greater depth 
than has yet been attained by the wells of this section. The 
floor of this valley has not, to my knowledge, been reached by 
any of the drill holes. The sheet of lava forming a portion of 
the southern margin of the valley is of later origin than some 
of the sedimentary strata and is evidently no part of the real 
floor of the valley. 

Fort Collins, Colorado. 



316 Andrews and Farr — Determination of Arsenic. 



Art. XXY. — The Volumetric Determination of Small 
Amounts of Arsenic; by Latjncelot W. Andrews arid 
Henry Y. Farr. 

The problem of the analytical determination of small amounts 
of arsenic, present as an impurity in products of the most varied 
character, is a constantly recurring one. The comparison of 
mirrors formed in the Marsh test in its several modifications 
and a few colorimetric processes have solved the problem for 
the special case in which the arsenic is present in traces only 
and in which a merely approximate estimation of the amount 
suffices. We have been led to seek a solution covering a wider 
range of cases, in which, for example, the arsenic may be 
present in any quantity from O'l mg., or possibly less, to 100 mg. 
and in which the accuracy of an ordinary quantitative analysis 
is required. 

Often, but not always, the distillation of the arsenic, after 
Fischer as arsenious chloride, furnishes satisfactory results 
under these conditions as a method of separation, but it 
always has the disadvantage of being rather time-consuming. 
It is better adapted for the determination of larger than for 
smaller amounts, since the distillate contains its arsenic in a 
highly dilute state and much contaminated with foreign salts, * 
or else in the form of arsenic acid,f which is not particularly 
well suited to quantitative determination. A method, in order 
to be perfectly satisfactory for the purpose in view, should 
comprise : (A) a process for the direct separation of the arsenic 
in a concentrated form, applicable to the greatest possible 
number of cases ; (B) a simple, rapid procedure for determin- 
ing the arsenic in the form in which it is furnished by step (A). 

It is obvious that none of the existing methods can meet 
these demands. The well-known Bettendorff test,;); in which 
the arsenic is precipitated in the metallic state by a mixture of 
hydrochloric acid and stannous chloride, appeared to us to pos- 
sess possibilities of development for the process we were look- 
ing for. This test has been used by E. S. Peck§ for the 
colorimetric determination of arsenic in metallic iron. Accord- 
ing to Bettendorff (1. c.) the sensitiveness of the reaction is 
such that it will show " most definitely " the presence, in one 
cubic centimeter, of two millionths of a milligram of ammo- 
nium-magnesium arsenate, that is, of less than a millionth 
milligram of metal. Our own observations on this point con- 

* Alkali chloride, if the distillate is received in alkali hydroxide solution. 
f When the distillate is received in nitric acid or other oxidizing agents. 
JFres. Zeitschr., ix, 105, 1869. 
§Pharm. Jour. (4), xiii, 130, 1901. 



Andrews and Farr — Determination of Arsenic. 317 

firm the statement quoted, provided, naturally, that a layer of 
sufficient thickness be examined and that the liquid is other- 
wise absolutely colorless. 

The arsenic as thrown down in the BettendorfT test, 
although finely divided, settles in two or three hours and can 
be filtered on the expiration of that time through an asbestos 
filter without exhibiting any tendency to run through, and can 
be washed with concentrated hydrochloric acid and with water, 
without undergoing oxidation. As pointed out by Betten- 
dorff, it contains more or less tin. Bettendorff found 95*86 to 
98*56 per cent of arsenic. The amount of tin, present as basic 
chlorides, varies, however, over a much wider range, than these 
figures would indicate, depending on the amount of oxidation 
undergone by the stannous chloride, the temperature, etc. The 
precipitate dissolves in decinormal, centinormal, or millinormal 
iodine solution, made slightly alkaline by addition of suitable 
amounts of sodium bicarbonate or phosphate. The readiness 
with which this solution occurs depends mainly on the amount 
of stannic acid or of basic stannic chlorides carried down with 
the precipitate. If the degree of this contamination is con- 
siderable, the precipitate must be shaken for several hours with 
the liquid in order to effect complete solution ; but if the 
amount be smaller or vanishing, solution takes place in less than 
a minute. This observation led us to prevent co-precipitation 
of basic tin compounds by adding tartaric acid to the reaction 
mixture of hydrochloric acid, stannous chloride and arsenite. 
There seems to be no inclination for the precipitate to entrain 
stannous compounds. 

Bettendorff found the reaction to be incomplete, even after 
the lapse of a long time, with hydrochloric acid of 23 per cent 
HC1, while with such of 25 per cent it was complete " after 
some minutes." This concentration represents the minimum. 
It is better to operate with a concentration of nearly 30 per 
cent, which presents no difficulties when the ordinary reagent 
acid of 40 per cent is used. That is to say, 25 cc of the latter 
acid should be employed as a minimum to each 10 cc of the 
solution to be examined.* The most obvious cause of errors 
lies in the possibility of loss by evaporation of the highly vol- 
atile arsenious chloride during the reduction process. This we 
guarded against in our earliest experiments by sealing up the 
reaction mixture in glass tubes, an effective but inconvenient 

* This rests of course on the assumption that the substance to be exam- 
ined contains nothing to neutralize the hydrochloric acid. Otherwise a far 
greater amount of acid may become necessary. Thus when tartar emetic is 
to be examined for arsenic, the salt is dissolved directly in the 40 per cent 
acid, and enough of the latter must be taken to furnish 25 to 30 per cent of 
absolute acid, after that consumed in converting the potassium and the anti- 
mony into chlorides has been duly allowed for. 



318 Andrews and Farr — Determination of Arsenic. 

expedient, which fortunately proved to be unnecessary, since 
sufficient protection is afforded by closing the solution in a 
bottle with a well-ground glass stopper, provided the temper- 
ature is not raised higher than 35° or 40°. The latter should 
not be lower than 25°, since the reaction then becomes unde- 
sirably slow. 

In the absence of tartaric acid, the arsenic does not usually 
visibly deposit on the glass walls, but when the precipitation 
of tin compounds is prevented, a small part of the arsenic 
sometimes forms an extremely thin film, adherent to the glass. 
The existence of such a film should be assumed, even when 
none is visible. Experiment has demonstrated that a minute 
truly, but titratable amount of arsenic may be present in this 
way on the seemingly quite clean surface of the vessel in 
which the reduction has taken place, 

The process finally assumed the following form. The mate- 
rial to be examined, if in the form of a solution more bulky 
than 20 cc . is neutralized and boiled down to 10 or 20 cc and 
transferred to a white bottle of 80 to 100 cc capacity, provided 
with a very well-ground stopper. To the liquid is added two 
and a half times its volume of the tin reagent, prepared by 
dissolving twenty grams of stannous chloride crystals and 
forty grams of tartaric acid in one liter of forty per cent hydro- 
chloric acid.* The stopper is inserted and the flask set aside 
to stand in a warm place till the precipitated arsenic has sub- 
sided, leaving the supernatant liquid water- whitef and clear. 
This usually takes about two or three hours, more or less, if 
the temperature is nearly 40°. An asbestos filter is prepared 
in the usual manner, either in a Gooch crucible or in a ISTeu- 
man filter tube. The precipitated arsenic is transferred to the 
filter with the aid of a small amount of concentrated hydro- 
chloric acid, rigorously chlorine-free. The flask is rinsed 
repeatedly with small amounts of water which are passed 
through the filter, which is completely washed by aid of suc- 
tion without allowing it to be exposed to the air more than 
necessary, except at the last draining, a precaution which may 
not be requisite, but which seems indicated. The proper 
amount of centinormal or of decinormal iodine solution is now 
measured into the flask by a pipette as indicated by the equa- 
tion, 

As + 51 + VNaHC0 3 = Na 2 HAs0 4 + 5NaI + 7C0 2 + 3ll 2 0, 
allowing for an excess of ten to one hundred per cent above 

* This reagent will remain colorless if the constituents are arsenic-free. 
It should be kept in bottles holding not more than 200 co and closed in such a 
manner as to give assurance of protection of the contents against oxidation. 

f Of course this does not apply when nickel chloride or other colored salts 
are present. 



Andrews and Farr — Determination of Arsenic. 319 

the theoretical.* The precipitate, with the asbestos filter, is 
now transferred quantitatively to the flask and shaken with the 
iodine solution. f Enough of a five per cent solution of sodium 
bicarbonate or of sodium phosphate should now be added to 
maintain neutrality throughout the reaction, but not an unduly 
large excess, and the shaking be continued till the asbestos is 
thoroughly disseminated through the liquid and till it is cer- 
tain that every particle of arsenic is dissolved. Fresh starch 
paste is added and the excess of iodine is titrated back by 
hundredth or thousandth normal arsenite solution. 

For quantities of arsenic smaller than 0*5 mg., thousandth 
normal solutions may be employed, but it must not be forgot- 
ten that in such high dilutions a correction must be employed 
for the amount of iodine required to produce the end-reaction, 
if satisfactorily exact results are required. This correction 
usually amounts to about 0*6 cc of N/1000 for each 50 cc of solu- 
tion, but it should be determined under the actual conditions 
of the titration. A main factor influencing the sensitiveness 
(aside from temperature) is the concentration of iodide in the 
liquid. For larger quantities than 10 mg. of arsenic up to 
100 mg. tenth normal iodine solution is preferably employed. 

The degree of precision to be expected in the results 
obtained by this method may be seen from the following test 
analyses. A hundredth normal arsenic solution (containing 
0'375 mg. As per cc.) was made by dissolving 990 - mg. of 
As 2 O s in water to make two liters, and also a fifth normal 

Table 





Arsenic 


Volume 


Vol. 












taken 


of 


of HC1 


Vol. of 


Vol. of 


Weight 






Milli- : 


solution 


used 


iodine sol. 


As sol. 


of As 


Error 




grams 


cc. 


cc. 


added 


req. 


found 


mg. 


1 


3-75 


10 


25 


50-02N/100 


25-02X100 


3*75 


o-oo 


2 


3-75 


10 


25 


41-05 


16-03 


3-753 


+ 0-003 


31 


3-75 


10 


50 


30-03 


5-23 


3-72 


— 0-03 


4- 


7-519 


20 


60 


60-06 


io-oo 


7-510 


— 001 


5 


0-375 


o 


10 


5-01 


2-46 


0-382 


+ 0007 


'3 


71-02 


10 


30 


51-OlN/lO 


35-2N/100 


71-23 


+ 0-21 


n 


75-00 


10 


50 


55-04N10 


49-4N/I00 


75-15 


+ 0-15 



^Xote. In analyses Xos. 3 and 7, abont 0*1 gram of crystallized copper 
sulphate was added, without, as may be seen, influencing the results. In 
another experiment, not recorded in the table, similar to No. 3 except that 
the copper was omitted, a solution of titanous chloride was substituted for 
the stannous chloride. In this case there was a negative error of 0'2 mg. in 
the arsenic. Twenty hours was required for the reaction, which is slower 
than when stannous chloride is used. 

* One cubic centimeter of centinormal iodine solution = - 15ing. As. 
+ It is ordinarily easy to see when the reaction is complete. When the 
amount of arsenic is less than 5 mg., it is practically instantaneous. 



320 Andrews and Farr — Determination of Arsenic. 

solution by dissolving 9*900 grams of the oxide in sodium 
hydroxide solution, saturating with carbon dioxide and then 
making up to one liter. The amounts of arsenic which appear 
in column 1 of the table were obtained by measuring the 
appropriate volume of one of these solutions. 

All volumetric apparatus employed was, of course, carefully 
standardized. The volumes of the standard solutions, as 
given in columns 4 and 5, are corrected for temperature and 
titre. 

A somewhat similar process to that described in the present 
paper has been published by Engel and Bernard.* These 
authors reduced the arsenic with a mixture of hypophosphor- 
ous and hydrochloric acids and titrated the precipitated metal 
in essentially the same manner as that adopted by us. Their 
test -analyses are very good on larger amounts of arsenic. 
They publish none for quantities less than 54 milligrams. 

The reaction is very much slower than it is with stannous 
chloride. This necessitates allowing the mixture to stand for 
twelve hours and then boiling, a proceeding which may well 
result in loss of arsenious chloride. 

It is far less easy to obtain hypophosphorous acid free from 
traces of arsenic than it is stannous chloride. The most 
important advantage which the use of the tin salt presents is 
probably to be found in the broad range of its applicability. 
In almost all of the salts which one ordinarily desires to sub- 
ject to an arsenic determination, no other operation is required. 
Even compounds of lead, bismuth or antimony need no pre- 
liminary separation. We have found that titanous chloride 
(TiCl 3 ) may be substituted for the tin salt, but without advan- 
tage so far as known. It is probable that the lower chlorides 
of chromium, molybdenum, or vanadium would answer the 
same purpose, but they hold out no promise of superiority. 

It is extremely likely that for the determination of fainter 
arsenical mirrors, obtained by the Marsh method, the iodo- 
metric titration will be found useful. A mirror of 0*01 mg. 
can scarcely be weighed satisfactorily, even with the best 
balance, because the weight of the tube when subjected to the 
action of the necessary reagents might change by an amount 
considerably in excess of the weight of the mirror. But it 
could be dissolved in millinormal iodine or starch iodidef and 
the excess titrated back by arsenite or thiosulphate of the 
same normality. For smaller mirrors than this, optical methods 
of comparison will still have to be used. 

Laboratories of the Mallinckrabt Chemical Works, 
St. Louis, Jan. 23, 1909. 

* Comptes rendus, cxxii, 390, 1896. 

f Compare Zeitschr. f. anorg. Chem., xxvi, 180, 1901. 



F. A. Per ret — Report on the Messina Earthquake. 321 



Art. XXVI. — Preliminary Pej>ort on the Messina Earth- 
quake of ' Deceniber 28, 1908 ; by Frank A. Ferret, K.I.C., 
former Honorary Assistant at Royal Yesuvian Observatory. 

As special representative of the American Consulate, the 
writer sailed from Naples on Dec. 30, arriving at the Straits 
at daybreak of the 31st.* He remained eight days and the 
scientific observations and photographs were chiefly confined 
to Messina and its environs, Reggio and Villa San Giovanni 
having been inspected only from the sea during a visit of the 
U.S. S. "Scorpion" to the Calabrian coast. The present 
report must, therefore, be considered as preliminary in its 
nature and limited by the extraordinary conditions incident to 
life in destroyed cities under martial law, aod in a state of 
siege. 

Before proceeding to a detailed account of the observations, 
it may be well to present a summary of the principal facts : 

For several weeks preceding the earthquake a number of 
more or less severe shocks were felt in the neighborhood of 
the Straits, the most important occurring on Nov. 5 and 
Dec. 10. Exactly twenty-four hours before the great event, 
i. e. at 5.20 a.m. of Dec. 27th, the seismograph at the Messina 
Observatory registered an important earth movement. 

Etna and Stromboli were unusually active on Dec. 25th, 
but neither showed sympathetic action at the time of the earth- 
quake nor immediately after. 

The earthquake occurred at 5.20 a. m. of Dec. 28, 1908 
(cf. fig. 1). The macroseismic duration was about 32 seconds. 

The epicentrum was apparently at the northern entrance of 
the Straits or a little to the E. and N. of this. 

The intensity within the raegaseismic area was between the 
9th and 10th grade of the Mercalli scale, and fell off rapidly 
with increasing distance from the epicentrum, indicating a 
centrum at no great depth, possibly 15 kilometers or less. 

The destructive area extends to and beyond Falmi on the 
north and to Ali on the south, say twenty miles in either 
direction. 

The isoseismals will show an elliptical form with the major 
axis lying N. and S. or, more precisely, from E. of N. to W. 
of S. 

Within the megaseismic area free surface waves were pro- 
duced, and their forms, preserved in the stone pavement of the 
Messina embankment, were photographed by the writer. 

* The cost of this seismologic study is being paid from a fund generously 
and promptly subscribed by friends of the Massachusetts Institute of Tech- 
nology and by the Volcanic Research Society of Springfield, Mass. 

Am. Jour. Sci. -Fourth Series, Vol. XXVII, No. 160.— April, 1909. 
22 



322 F. A. Ferret — Report on the Messina Earthquake. 

The event occurred on a steep barometric gradient, the mer- 
cury rising 5 mm during the night to a maximum at 9.00 a. m. 
of the 28th and then falling 10 mm in the course of the day. 

The moon was nearing its quadrature position of the 29th 
and had just passed its perigee, these conditions producing the 

Fig. 1. 




Fig. 1. Office of the Hotel Tiinacria, clock stopped at 5.23. 

u Terrestrial Maximum" of Dec. 28th on the writer's astro- 
seismic curve for 1908.* This combination was preceded by 
three very favorable luni-solar positions during the month of 
December, and this fact had led the writer to expect some 
* Science, Aug. 28, 1908. 



F. A. Perret — Report on the Messina Earthquake. 323 

important volcanic or seismic event before the close of the 
year. 

At the moment of the earthquake the moon was near the 
nadir and the sun just below the eastern horizon. 

The earth movement resulted in a sea wave which arrived 
at Messina two or three minutes after the shock and at Villa 
San Giovanni several minutes later. It reached the point of 
Schiso (^Naxos) in 35 minutes and Malta in 115 minutes. It 
was also registered by the mareograph at Ischia, the greatest 
rise (22 cm ) occurring at 2.30 p.m. and the intervals between 
crests being 12 minutes. The height of the wave at Messina 

Fig. 2. 




** *.:.:W ' fU * 






XM:r * * 




Fig. 2. View along water-front, showing relative stability of low buildings. 



seems not to have exceeded three meters ; at Keggio it was 
somewhat greater and reached its maximum on the coast below 
Taormina. In all cases the wave was noteworthy only by its 
development on reaching shallow water — a small, low free- 
board ferry-boat with passengers aboard having been in the 
Straits at the time and suffering no inconvenience beyond the 
difficulty of landing at the damaged and submerged ferry- 
slips. 

The earthquake is described as having been preceded for a 
few seconds by a singing sound like a far-away wind storm 
which rapidly drew nearer and became a rumble and a roar 



when the earth movement b 



esan. 



321 



Report on the Messina Earthquake. 



Nearly all the after-shocks observed by the writer were 
accompanied by sound phenomena. 

Observers at Taormina report having seen luminous effects 
on the horizon in the direction of Messina immediately before 
the earthquake, but at Messina all was dark. The sea appeared 
luminous, possibly because of having been converted into foam 
by the vibrations. There is no doubt of luminous effects hav- 
ing been observed immediately before and during other earth- 
quakes in this region — that of 1905 being accompanied by a 
strong red glow upon the mountain tops. 

Fig. 3. 




Fig. 3. Collapse of the embankment. 

The Messina seismograph recorded a part of the movement. 

The loss of life and property was enormous owing to the 
"rubble" construction of the buildings. 

As has often been noted, those buildings which presented 
their diagonal to the direction of the earth movement resisted 
better than those whose faces paralleled or right angled the 
motion, and those on loose or sloping ground suffered most. 

In Messina the general direction or " throw " of the move- 
ment was from JST.E. to S.W., and the same was the case at 
Villa San Giovanni. 

There seems to be no evidence of a geological sinking. The 
sea has advanced in places as much as one hundred meters, but 



F. A. Perret — Report on the Messina Earthquake. 325 

this is evidently due to the downslip of loose material into deep 
water. The Messina embankment, built of lava blocks on 
made ground, has collapsed under the influence of the earth 
vibration and the sea wave, but the relative insignificance of 
its fall is shown by the perfectly upright walls of houses not 
twenty feet away. 

Unless subsequent soundings shall demonstrate the contrary, 
it does not seem that any great physical changes have taken 



Fig. 4. 




Fig. 4. Partial collapse of the embankment. 



place upon the earth's surface in consequence of the earth- 
quake. That of 1783 resulted in considerable downfalls along 
mountain ranges, the formation of lakes by the sinking of 
plains and the production of crevasses. In the present case 
these latter are limited to insignificant fissures. 

The earthquake was lightly felt at Naples and more strongly 
on the island of Capri. 



326 F. A. Per ret — Report on the Messina Earthquake. 

Properly constructed houses in Calabria withstood the shock. 

Entering the Straits at daybreak of the 31st, I observed 
that the famous rock of Scylla, which, together with the light- 
house (Faro di Messina) on the opposite point and several of 
the Lipari Islands, was reported as having disappeared, stood 
in its accustomed place. The lighthouse also was standing, 
but the lantern at the top had been slightly displaced toward 
the east. The first view of Messina from the sea did not give 

Fig. 5. 





■ ; -'.. ■■■■£■". §'% • :fi 


f"' '*"":; '-^j^;, ., iff 

1 ^ 111 1 


|^;;ii'« r | : ;; 


'. ' i: " 




"^3 : Kf 11 


:'..:. ■■■"■■..■■■> 


li^ 


,';"'■: 




. . ■ ■ . ..:..■.■ .■■:■■: ■ . ■ ■ ■ ...■■■ 


|t:^|||: 


■ ••€£*** >: 




Is 




^^^ 






Pe 











Fig. 5. Free surface wave-forms retained by curb stones of the embankment. 

the impression of a complete disaster. A considerable portion 
of the facades of the line of buildings along the quay — the 
famous " Palazzata '' — remained standing as well as a num- 
ber of one or two-storied houses, but a nearer view showed 
that there was practically nothing back of the facades and that 
these houses were damaged, although still standing (fig. 2). 

On going ashore it was easily seen that the embankment had 
collapsed from the sliding down into the deep harbor of its 
insecure foundations (tigs. 3, 4). All along the quay a critical 
eye could detect signs of the free surface wave, and in one spot 
the curb had become detached from the sidewalk and retained 
admirably the full wave form into which it had been thrown. 
(See fig. 5.) The average distance from crest to crest of these 
waves was two meters and the height from trough to crest 



F. A. Ferret — Report on the Messina Earthquake. 327 

(double amplitude) was 30 cm . In other places the wave was of 
natter form, i. e., height 16 cm , crest to crest 6 m. 

The " rubble " construction of the buildings was apparent all 
along the quay, broken arches of the '* Palazzata " revealing 
the small round stones set in mortar of poor quality (figs. 6, 7). 



Fig. 





^ , 


ft^iifc 


; \ ,_, ' "f-.. 


. 


"""?" ""' •';4^^^t. M ; •-■ •■•-■'•! 


,**~ ~ 


% : - ^ : ; ■■■■■' ; - *; 4 


• 


• ' : ■ i-?. ; u„- . \ . ~ ,~ .» ■ 




******** ' » m 

m 1 


v 


W£%**. 3»1 




mm ■■:*■;■:'■;■ 


w 


Wm 1 


M> J- 


;:; ; *.-: ■ 




iHllH 
z^Wlmmw.] 


1 


.^ 


" 


m Sf Mf^ 




' 










;-3^i^u^' i; ; .&38S&^. .'...) 


'*Y LXJ W^Ha^^ 




..- ■ ■ 


■„ ; .«:& :■■■.:.: C;-'l-, ' 



Fig. 6. Showing rubble construction. 



A conspicuous exception was seen in the building occupied by 
the French Consulate, the walls of which were fairly thick 
and built of tiling or thin brick, its chief defect being the 
weakness of the floors and its height of four stories (fig. 8). 
The relative stability of these walls was shown by a seven-foot 
pier glass which was not even cracked. The American Con- 
sulate building had collapsed to a compact heap of mortar and 
stone. 



328 F. A. Perret — Report on the Messina Earthquake. 

The general direction of the earth movement was well shown 
by the wreck of the Manrolico monument in the Yilla 
Mazzini (fig. 9). This has fallen due S.W. and, having been 
symmetrical and standing on a level plain, it gave as fair an 
index as could be obtained. The same general direction was 



Fig. 




Fig. 7. Broken arch of the Palazzata showing rubble construction. 



observed just outside the city, where the top of a tall stack had 
been snapped off and fallen S.S.W. (fig. 10). Two other single 
column monuments on the hills west of the city had also fallen 
S.W. At Yilla San Giovanni a tall stack was tilted S.W. 
Although in Japan it is generally found that such objects fall 
inward or toward the epicentrum, this was not the case here, 
the most probable explanation being that, owing to the slight 
depth of the centrum and the nearness of the city to the epi- 



F. A. Perret — Report on the Messina Earthquake. 329 

centrum, there was a very pronounced vertical component at 
the beginning of the earth movement which, in combination 
with the horizontal motion, threw the objects in the direction 
of the movement instead of tripping their bases and causing 
them to fall inward. This will be more evident to the reader 



Fig. 8. 




Fig. 8. Euin of French Consulate. Walls fairly thick and well built but 
floors weak and house too high for this type of construction. 

from an inspection of the photograph of the Maurolico monu- 
ment, the dome of which has been thrown to a distance of five 
meters from the center, which would scarcely have been the 
case if it had fallen under the impact of horizontal motion. 
The lantern of the Faro forms a case of displacement toward 
the epicentrum. 

The principal after-shocks observed by the writer were the 
following : 



330 F. A. Ferret — Report on the Messina Earthquake. 

Jan. 2, — 8.14 a. m. — f sec, weak. 

9.40 p. m. — \ sec, strong, threw down walls. 
11.42 p. m. — two weak shocks \ sec apart. 
Jan. 3, — 3.55 a. m. ) 

4.56 a. m. y weak, brief duration. 
7.24 a. m. ) 
Jan. 5, — 12.10 noon — f sec, strong ; several weak shocks two to 
three minutes later. 



Jan. 



5.05 p. m. 

9.50 P. M. 

9.00 A. M. 

10.00 A. M. 



unimportant, 
unimportant. 
Fig. 9. 




Fig. 9. The Maurolico Monument in the Villa Mazzini ; this has fallen due S.W. 

Jan. 7, — 5.00 a. m. — strong, brief duration. 

6.28 p. m. — 5 sec, strong, threw down walls, followed by 
a replica. 
Several shocks during the night and the early 
morning of Jan. 8th. 



This should not be taken as 



occurring during the time 



complete record of all shocks 
nor can the relative intensity be 
depended upon as accurate, the various observations having 
been made on shore, on shipboard and in the ruins. The need 
of a good, portable seismograph was never better illustrated 
than by this event, as all the instruments in the neighbourhood 
w r ere destroyed. It would have been possible to take a portable 



F. A. Per ret— Report on the Messina Earthquake. 331 

instrument down to Messina from Naples and to have thus 
obtained a complete record of all after-shocks. 

Of those indicated above three are deserving of special 
mention. That of Jan. 2, at 9.40 p. m., occurred when the writer 
was standing on the deck of a steamer moored to the embank- 
ment. The impression was that of a submarine explosion — a 
loud report and a sharp vertical movement. These, however, 

Fig. 10. 




Fig. 10. Tall stack ; top snapped off and fallen S.S.W. 



were partly due to the iron hull of the ship receiving the 
impact from the water. In from three to live seconds was 
heard the crash of falling walls within the cit}^ and smouldering 
tires blazed up anew. This shock is reported to have snapped 
the anchor tips of the British cruiser 
carried two miles to the southward by 
through the Straits. 

At ten minutes past noon of the 5th another strong shock 
of the same general nature occurred, but this was felt on shore 



" Exmouth," which was 
a twelve-knot current 



332 F. A. Perret — Report on the Messina Earthquake. 

only as a horizontal wave motion. It was followed by several 
weaker movements. 

At 6.28 p. m. of Jan. 7, a shock lasting fully five seconds 
formed the most interesting of those observed. The duration 
was such as to give ample time to study the phenomenon and 
it was impossible to avoid the conviction that the originating 
movement at the centrum had a duration not greatly inferior 
to that of the observed effect. This is contrary to the accepted 
ideas of the day regarding earthquake generation and a dis- 
cussion of the subject may be reserved for a future paper, but 

Fig. 11. 




Fig-. 11. Fronts fallen from houses, lying East and West. 

the writer feels in duty bound to record the impression in the 
belief that no honest observation is without value. 

This shock was experienced when on board the U. S. S. 
" Scorpion " and the man on watch reported having " seen " 
the earthquake pass through the city from N. to S. When 
interrogated he could give no more definite information, but 
it is evident that the earth waves had produced a visible 
undulation of the buildings and walls along the water front. 
Many walls fell and the shock was followed by another not as 
strong. A number of shocks occurred during the night and 
in the early morning of the 8th. 

As to the cause of these Calabrian earthquakes, the writer 
inclines to the opinion of Mercalli, viz.: that they are due to 
the movements of deep-seated magma and belong, therefore, 



F. A. Ferret — Report on the Messina Earthquake. 333 

to the type which he denominates " inter-volcanic." In their 
nature they are, of course, tectonic, and I often permit myself 
to ask if the primal cause of all tectonic earthquakes may not 
yet be found in magmatic intrusion, the fact of their non- 
occurrence in the immediate neighborhood of active volcanic 
vents and of their prevalence in the steeply folded portions of 
the earth's surface constituting, in my opinion, an argument 
for, and not against, the hypothesis. 

Fig. 12. 




Fig. 12. North end of Quay. 



At all events, this portion of the Italian peninsula lying, as 
it does, between the Tyrrhenian and Ionian deeps and subject 
to the upheaval revealed by the Quaternary terraces of the 
Aspromonte, must be considered as one of the most pro- 
nouncedly seismic areas of the globe. This being the case, it is 
idle and harmful to encourage the hope that this region will not 
be subject in the future, as it has been in the past, to frequent 
and severe earthquakes. Rather should it be impressed upon 
both government and people that, sooner or later, these are 
certain to occur and that the proper construction of houses to 
withstand their effects is an absolute necessity. Only thus, 
with the active prosecution of the study of prediction, may we 



3ott F. A. Per ret — Bejjort on the Messina Earthquake. 

hope to avoid future repetitious of the recent great disaster, 
and ample means should, therefore, be provided for putting 
earth-science on a world-wide basis and bringing it thus to at 

Fig. 13 




Fig. 13. Wreck of a tall, poorly constructed building. 



least a par with astronomy. A glance at a list of the volcanic 
and seismic catastrophes of the last eighteen years will suffice 
to show that few other lines of scientific investigation can vie 
with this in importance to the human race. 
Vomero, Naples, Italy. 



Maury — Connecting Link in the Genesis of Fulgur. 335 



Art. XXYII. — A New Connecting Link in the Genesis of 
Fulgur y by Carlotta J. Maury. 



Among a quantity of fossils lately collected by Professor 
G. D. Harris at Montgomery, Louisiana, from the Jackson 
horizon of the Eocene is a very interesting species which forms 
a perfect connecting link between Levifusus and Fulgur. A 
dozen specimens were found. 

The relationship of the two genera was pointed out some 
years ago by Dr. Dall and Professor Harris. Dr. Dall in 1890 
stated that Fulgur, which took its rise in the Eocene, was 
descended from such forms as Levifusus Blakei and trabeatus. 
Professor Harris later noted the tendency shown by many 
Levifusi to revert to an ancestral Fleurotoma-Uke form and 
traced the derivation of Fulgur from 
Pleurotoma through Surcula, Levifusus 
(pagoda-like forms), Levifusus (suteri- 
j i k e forms), Levifusus {trabeatus-\ike 
forms), Levifusus Branneri to Fulgur 
echinaium. 

But Levifusus Branneri Harris in addi- 
tion to shoulder spines is ornamented by 
a row of twelve nodules on the center of 
the body whorl. The Montgomery shell 
is without the slightest trace of this row, 
and, as shown in the accompanying figure, 
presents such a striking resemblance to 
Fulgur that it might almost be taken as 
one of the many varietal forms of Fulgur 
spiniger. The spire, however, is almost 
exactly that of Levifusus Branneri. To emphasize the fact 
that it is the most direct ancestor of Fulgur known, the name 
fulguriparens has been given to the Montgomery shell. 




Levifusus fulgnriparens 
n. s. (Longitude 25 mm .) 



Hastings-on-Hudson, N. Y. 



336 Scientific Intelligence. 

SCIENTIFIC INTELLIGENCE. 

I. Chemistry and Physics. 

1. A Method for Calculating the Boiling 'joints of Metals. — 
The view was advanced by F. Krafft about four years ago that 
the essential process of boiling in a vacuum consists in overcom- 
ing gravity, and that this, on the surface of the earth, is equiva- 
lent to the atmospheric pressure. As a proof of this fact he 
showed that the elements mercury, cadmium, zinc, potassium, 
sodium, bismuth and silver require just the same addition of heat 
to bring them from the point of the commencement of volatiliza- 
tion in a vacuum to the point of boiling in a vacuum, as is 
required to bring them from the latter point to the point of boil- 
ing under atmospheric pressure. In order to obtain further 
data in regard to this interesting relation, Krafft and Knocke 
have determined the data under consideration for arsenic, and 
have found : 

Commencement of evaporation, ,nm . 96° 

Sublimes at mra 325° 

Sublimes at 760 m,n 554° 

Difference in each case ... 229° 

They found with thallium : 

Commencement of evaporation, mm 174° 

Boiling-point at mm ... 818° 

Since the difference here is 644°, the calculated boiling point at 
760 mm is 1462°. In this manner, from data furnished by Krafft 
and his co-workers, Moissan calculated the approximate boiling 
point of copper as 2240° and of gold as 2530°, and he found by 
distilling in an electric furnace an alloy consisting of equal parts 
of copper and gold, that the copper boiled off faster than the 
gold, thus showing a lower boiling point of the copper, as indi- 
cated by the calculation. — Berichte, xlii, 202. h. l. w. 

2. Some Properties of the Radium Emanation. — About two 
years ago Rutherford observed that the emanations of radium, 
actinium, and thorium were completely absorbed by cocoanut 
charcoal at ordinary temperatures. He has recently repeated 
this experiment with the radium emanation, using much larger 
quantities, and has found that the actual volume of this emanation 
capable of absorption at room temperatures is very small. For 
example, several grams of the charcoal are required to absorb 
completely the emanation from 200 mg. of radium at ordinary 
temperature, although the volume of the gas is only one-tenth of 
a cubic millimeter. As was to be expected, the absorptive 
power of the charcoal increases rapidly with the lowering of the 
temperature. It was found that 0*8g. of charcoal from which 



Chemistry and Physics. 337 

air had been removed by heating absorbed the emanation from 
83 mg. of radium (about 0-05 cubic mm.) at —150° C. As the 
temperature was slowly raised, less than one-tenth of the 
absorbed emanation could be pumped out at — 50° C, as was 
shown by measurement by the y-ray method. Above —40° C. 
the emanation commenced to escape rapidly, and half had been 
pumped off at 10° C. About 10 per cent remained at 100° C, but 
practically all was released at the temperature of the softening 
of glass. The results show that at 10° C. the charcoal absorbs 
about 0*03 cubic mm. per gram, and at — 40° C about 0*06 cubic 
mm per gram. — Chem. JVeios, xcix, 76. h. l. w. 

3. Method for Preparing Hydrogen Phosphide. — Matignon 
and Teannoy have devised a simplified method for the prepara- 
tion of the gas PH 3 . They mix powdered calcium phosphate 
which has been previously calcined, in order to remove every 
trace of moisture with powdered aluminium, in the proportions 
required by the equation 

3Ca 3 (P0 4 ) 2 + 8Al 2 = 3Ca 3 P 2 + 8Al 2 3 

put the mixture in a crucible, heat it to dull redness, and ignite 
it with an ignition mixture according to the Goldschmidt method. 
The products of the reduction do not separate, but form a brownish 
mixture, which after being broken up is suitable for the produc- 
tion of hydrogen phosphide, by the action of water, followed by 
hydrochloric acid towards the end of the operation. The result- 
ing gas was found to contain 2 or 3 per cent, of hydrogen, but 
otherwise it appeared to be perfectly pure.— Comptes Rendus y 
cxlviii, 16V. h. l. w. 

4. The Theory of Valency, by J. Newton Friend. 12 mo, 
pp. 180. London, 1909 (Longmans, Green and Co.). — This work 
is one of a series of text-books of Physical Chemistry, edited 
by Sir William Ramsay. Nine of these monographs have 
already appeared, and five more are announced as in the course of 
preparation. The book under consideration gives a concise 
account of the more important theories of chemical combinations 
w 7 hich have exercised the minds of scientific men down to the 
present day. It appears that heretofore there has been no 
treatise in the English language upon the important subject of 
valency, while in German there is only one, a not very exhaustive 
work which was published several years ago. The book will be 
useful to students of chemical theory, since it covers the ground 
very thoroughly and gives a very full list of references to the 
literature. h. l. w. 

5. Recent Advances in Organic Chemistry, by A. W. 
Stewart. 8vo, pp. xv, 296. London, 1908 (Longmans, Green 
and Co.). — The author has endeavored to give an idea of some of 
the most recent researches in organic chemistry which have been 
carried out within the last ten years. The subject has been con- 
sidered from a synthetic point of view and the work has some 
commendable features, but is too condensed to be of much value 

Am. Jour. Sci.— Fourth Series, Vol. XXVII, No. 160.— April, 1909. 
23 



338 Scientific Intelligence. 

to one who is familiar with the advances made in recent years. 
While many important researches unfortunately have not been 
considered, nevertheless the subject has been treated in such a 
manner that the book will tend to stimulate the enthusiasm of 
students of organic chemistry. t. b. j. 

6. Michelson's Ether Research. — Emil Kohl, after a close 
scrutiny of Michelson's celebrated experiment, concludes that a 
careful study of the formation of the interference lines which are 
the essential feature of this experiment is necessary before a con- 
clusion can be reached in regard to the negative results. He calls 
attention to the importance of the factor of the distance; and 
discusses the questions which arise if ordinary light is emploj^ed 
instead of monochromatic, and recommends that the experiments 
should be repeated with attention to these points. — Ann. der 
Physik, No. 2, 1909, pp. 259-307. j t. 

7. Influence of Pressure upon Thermoelectric Force. — Hein- 
rich Horig employed a pressure of 1400 kg /cm -2 and found 
that the thermoelectric force of a platinum mercury combination, 
which before pressure at M=. 150°, gave an electromotive force 
of 10 _c volt per degree increase of temperature. When this 
thermo-element was submitted under the same conditions of tem- 
perature to a pressure — it gave for kg /cm -2 a change per degree 
At of 2*1 8 '10 -10 volt. The current direction of this pressure 
effect was opposed to the original thermoelectric effect, and up to 
a pressure of 1400 kg /cm -2 no deviation from proportionality 
with the pressure was observed. Similar results were obtained 
with a eutectic KNa alloy. — Ann. der Physik, No. 2, 1909, pp. 
371-412. j. t. 

8. Aluminum Cell as a Condenser. — The use of this property 
of aluminum is attracting much attention in view of its possible 
practical use with alternating currents. J. de Modzelewskt 
carried out experiments with a cell having aluminum anodes and 
Ni cathode in a six per cent ammonium bicarbonate solution and 
found that slow and gradual increase of voltage was necessary 
for a high degree of formation. The capacity of the condenser 
was about 15 mfd. The author indicates the need of prelimi- 
nary forming at a low voltage. It is said that temperature plays 
an important part in* the formation. The author does not refer 
to this fact. — Lumiere Electr., Aug. 8, 1908, pp. 187-188. j. t. 

9. Changes in the Spectra of Gases submitted to the Magnetic 
Field. — A. Dufotjr has studied the Zeeman effect in a number 
of rarified gases, and gives photographs of the normal and abnor- 
mal effects observed. He believes that the results obtained con- 
firm the theory of the presence of negative and positive electrons. 
—Physik. Zeitschrift, Jan. 15, 1909, pp. 124-138. j. t. 

10. Die elektrischen Eigenschaften und die Bedeutung des 
Selens fur die Elektrotechnik • von. Chr. Ries. Pp. 93, 52 
figures. Berlin, 1908 (Administration "Der Mechaniker" ).— 
This volume of nearly 500 pages gives an interesting account of 
the properties of selenium, particularly as regards the connection 



Geology and Natural History. 339 

between the temperature and. light and its electrical conductivity. 
This is a subject to which much attention has been given in the 
last decade, and numerous applications in the arts have been 
found, the most important of which are the light telephone and 
electrical tele-photography. These technical uses of selenium are 
well described and illustrated. 

1 1 . Physics for Secondary Schools ; by Charles F. Adams. 
Pp. 490. New York, 1908 (American Book Co.). Elements of 
Physics ; by George A. Hoadley. Pp. 464. New York, 1908 
(American Book Co.). — The above are new text-books, alike in 
their design to meet the requirements of College Entrance Boards 
and the various State and Association reports. They are interest- 
ingly written, are well illustrated and contain a large number of 
problems. Each contains an Appendix : that of the former 
including a discussion of variation and proportion, a table of 
natural sines and tangents and a table of constants for copper 
wire ; that of the latter, answers to problems and. a summary of 
the fundamental formula?. d. a. k. 



II. Geology and Natural History. 

1. Iowa Geological Survey, Annual Report for 1907 y 
Samuel Calvin, State Geologist. Vol. xviii, pp. 386, 16 plates, 
41 figures. Des Moines, 1908. — The Iowa Surve}^ is continuing 
its researches in the coal resources of the state and has undertaken 
investigations of the location, extent, and character of the peat 
deposits. Arrangements have also been made with the U. S. 
Geological Survey for cooperation in securing a topographic map 
for the entire state. In addition to the statistics of mineral 
production, the report for this year contains an important paper 
by Charles R. Eastman, on the Devonian Fishes of Iowa, treated 
under the following heads : Relations of Paleoichthyology to 
Biology ; Stratigraphy of the Devonian Fish-bearing Beds of 
Iowa; Evolutionary History of Fishes; Systematic account of 
Devonian Agnatha and Pisces including the subclasses Elasmo- 
branchii, Holocephali, Dipneusti, Teleostomi ; Faunal lists. 
Three maps by Professor Schuchert showing the Devonian 
paleogeography are included in the illustrations. h. e. g. 

2. Oklahoma Geological Survey ; Chas. N. Gould, Director. 
Bulletin No. 1, Preliminary Report on the Mineral Resources of 
Oklahoma; by Chas. N. Gould, L. L. Hutchison and Gaylord 
Nelson. 80 pp., 11 figs. Norman, 1908. — The recently founded 
Oklahoma Survey is justifying its existence by a timely study of 
the economic resources of the state. Among other papers, the 
first volume includes preliminary reports on coal, oil and gas, 
asphalt, gypsum, salt, lead and zinc, glass sand, and building- 
stones, including clay and Portland cement. h. e. g. 



340 Scientific Intelligence. 

3. Glaciation of the Uinta and Wasatch Mountains ; by 
Wallace W. Atwood. U. S. Geological Survey, Professional 
Paper No. 61. Pp. 93, 15 plates, 24 figures. Washington, 1909.-— 
" Every large canyon that heads near the crest of the range (Uinta) 
has been glaciated," and evidence of at least two epochs of glacia- 
tion appears. During the first epoch the lower limits of the ice 
were 8,000 feet on the north side of the range and 7,000 feet on 
the south ; for the later epoch the corresponding figures are 
8,400 and 8,000 feet. The maximum extent of glaciation in an 
east-west direction was 82 miles, in a north-south direction 42 
miles and within this area were one hundred and four glaciers over 
a mile in length, eight of which were over 20 miles long, the 
longest extending 27^ miles. 

In the Wasatch range there were 50 Pleistocene glaciers exceed- 
ing a mile in length, extending on the east side of the range to an. 
altitude as low as 6,000 to 7,000 feet; on the west side 14 glaciers 
reached an altitude of less than 6,000 feet and 7 of the 50 larger 
glaciers reached the shores of Bonneville and parts of their 
moraines are buried by lake deposits. A very interesting feature 
of the work is the proof of at least two distinct glacial epochs 
in the Wasatch, corresponding with the periods of humidity in 
old Lake Bonneville. 

Mr. Atwood's paper is an important contribution to glacial 
geology and has given a definite quantitative form to the theories 
and surmises of the 40th Parallel Survey. h. e. g. 

4. Glacial Waters in Central New York ; by H. L. 
Fairchild. New York State Museum, Bull. 127, 1909. Pp. 61, 
42 plates. Albany. — Previous papers by Professor Fairchild 
have dealt in great detail with the Pleistocene history of central 
New York. The present paper deals with the ice-border drain- 
age, particularly with reference to its connection with the present 
water bodies of this district. The tracing of the ice work in this 
region necessarily involves detailed descriptions of local phe- 
nomena, which serve as types of similar occurrences elsewhere. 
The glacial lake succession in New York state is well brought 
out by a series of maps which trace the water bodies through the 
following stages : Local glacial lakes, Lake Newberry, Lake Hall, 
Lake Vanuxem, free drainage, Lake Warren, Lake Dana, Lake 
Dawson, Lake Iroquois. The maps and illustrations are up to the 
high standard set by the New York Survey. h. e. a. 

5. Ground Waters of the India Region, California / by 
Walter C. Mendenhall. IT. S. Geological Survey, Professional 
Paper 225. Pp. 53, 12 plates, 5 figures. Washington, 1909.— 
Included in this report on the water resources of part of the 
desert region of southern California is a geographic and geologic 
sketch of the Colorado desert, a region made known by the 
explorations of William P. Blake in 1853, since which time it 
has been scarcely at all studied by geologists. Mr. Mendenhall 
shows that the desert lowland, including the Gulf of Mexico, is 
due to faulting, — part of it since the Tertiary beds were depos- 



Geology and Natural History. 341 

ited, — and that Santa Rosa ridge has the topographic character- 
istics of a faulted block. The structure of the valley floor has 
been determined and a study made of the origin and character 
of the delta silts, consolidated Tertiary beds, sand dunes and 
saline deposits. h. e. g. 

6. Die Alpen im JElszeit alter ; von Dr. Albreoht Penck 
und Dr. Eduard Bruckner. In three vols., 1 1 T6 pp., 37 plates, 
136 figures, 19 maps. Leipzig (Tauchnitz). — With the publication 
of parts 9, 10, and 11, including the continuation of the study of 
the glaciation of the southern and eastern Alps, this great work is 
brought to its close. The concluding parts on Glacial Physiog- 
raphy (pp. 1141-1152) and the chronology of the Ice Age (pp. 
1153-1156) by Professor Penck present the prominent results of 
glaciation in a clear manner and make it possible to compare the 
glacial history of the Alps with that of other regions. 

The first instalment of this work was issued in December, 1901, 
and the concluding chapter bears the date of December, 1908. 
While much of the field work on which it is based had been 
previously done, new observations and development of new 
theories have made it necessary to enlarge the original scope and 
to delay, accordingly, the final completion of the work. As it 
stands, it is probably the most important work on glacial geology 
ever issued, not only for the detailed description and explanation 
of this classic region for glacial study, but also for its contribu- 
tions to the general science of glaciology. h. e. g. 

7. Geological Survey of Western Australia; A. Gibb 
Maitlaxd, Government Geologist, 1908. Perth. — Two bulletins 
have recently been issued by this organization, viz. : Bulletin 31, 
Part I. — The Bonmevale and Kunanalling Districts, Coolgardie 
Goldfield ; by Chas. G. Gibson, Assistant Geologist. Pp. 56 
with 6 plates, 2 figures, 3 maps. Part IT, The Black Range 
District, East Murchison Goldfield. Pp. 65-116, 3 plates, 3 maps. 
Bulletin No. 34, — Report upon the Auriferous Deposits of 
Barrambie and Errolls (Cue District) and Gum Creek (Nannine 
District) in the Murchison Goldfield ; also Wiluna (Lawlers 
District) in the East Murchison Goldfield ; by Chas. G. Gibson. 
Pp. 40, 3 plates, 3 maps, 6 photographs. 

8. Die Geologischen Grundlagen cler Absta miming slehre; von 
Gustav St ei x:\i a xx. Pp. viii, 284. Leipzig, 1908 (W. Engelmann). 
— The author is in the main reactionary and out of harmony with 
the methods and explanation usually given as to the causes for 
organic change. Lamarck and Goethe are his standard-bearers. 
The book is dedicated to Lamarck and from his work he selects 
as his text "All living families subsist in spite of their variation." 
The historical method or the appearance of organisms in sequence 
of time plus Steinmann are his main principles. Darwin's book, 
he states, should not have been called " Origin of Species " but 
" Changeability of Organisms." Through paleontology alone is 
it possible to understand the changes that have taken place in the 
organic world. To ask the object of an organ or an organization 



342 Scientific Intelligence. 

and to attempt to explain the degree of their usefulness is unsci- 
entific. 

The problems in the development of life are : (l) Disappear- 
ance of extensive groups of plants and animals ; (2) sudden and 
extensive development of new groups ; (3) absence of connecting 
forms between great groups of plants and animals, and (4) lack 
of understanding of the entire development of life. The first 
problem seemingly has its explanation in the inter-changing of 
sea and land and the resulting climatic variations ; paleontologic 
biota, on the other hand, are but fragments of former progressive 
and repressive assemblages. 2. Imperfection of the local and 
regional paleontologic records is more apparent than real ; present 
known record not only a fragmentary but very much scattered 
one ; new invading biota indicate lost records, elsewhere recover- 
able, therefore no actual sudden appearance of new groups. 3. 
Phylogenetic material as yet insufficient and too greatly scattered 
in museums ; of small groups there is much connecting material 
but none between classes, orders, etc. 4. Present methods of 
explanation not scientific and according to nature but philosophic. 

Stein mann's phylogenetic-historic methods are peculiarly his 
own. He complains (wrongfully) about systematists basing their 
classification on single characters, thereby displacing the true 
relationships of forms. He would take the entire organism into 
consideration and group them historically. We will test his 
results along a single line, namely the Brachiopoda. The family 
Orthidae is said to be extinct since the Permian, but Steinmann 
would have us believe that descendants of this stock in Dahna- 
nella still live in Megerlea and Kraussina. His progressive 
series are the orthid Dalmanella, the rhynchonellid Rhynchonel- 
lina, the terebratuiids Megerlea and Kraussina, because all are 
said to have the same general external expression (homomorphic) 
while the arm supports are progressive from crura to loops. 
Testing this sequence in the light of chronology, we see that Dal- 
manella disappears with the Devonian while Rhynchonella does 
not appear until, and is restricted to, the Jurassic, and the other 
genera do not appear until the Tertiary. No attention is paid to 
the fact that Rhynchonelllna is impunctate, while the punctate 
shell structure of Megerlea and Kraussina is very different from 
Dalmanella. Then too no orthid has deltidial plates while the 
development of the loop both ontogenetically and chronogeneti- 
cally in Megerlea and Kraussina indicates beyond a doubt that 
these genera never arose in Rhynchonellina and not at all in 
Dalmanella. 

The cemented bivalves Rudistidae are said to have the same 
general organization as the ascidians and the slight differences 
that exist are thought to be due to reduction in the latter, caused 
by prolonged (geological) sessility. Since Cretaceous times the 
Rudistid descendants have lost their shells and are now the 
ascidians. On the other hand, Salpa are shell-less brachiopods 



Geology and Natter at History. 343 

that have originated in Productus through JRichthofenia, a highly 
modified sessile productid. 

Does progress lie along the path blazed by Steinmann ? 

c. s. 

9. Miner alien- Sammlung en. Ein Hand- nnd Hilfsbuch filr 
Anlage und lnstandhaltung Miner alogischer iSammlungen; von 
Wolfgang Prendler. I Teil. Pp. viii, 220, with 314 figures. 
Leipzig, 1908 (VV. Engelmann). — This is the first part of a work 
which is planned to aid those who are concerned with the instal- 
lation of mineral collections. It gives, in connection with some 
practical instructions in regard to the collection of minerals, a 
clear and coucise summary of the crystallographic and physical 
characters of crystals. A novel feature is the discussion of the 
arrangement of the collections, as regards cases, methods of 
arrangement, labeling, etc., which gives many useful hints not 
often found in print. 

10. Jadeite from Upioer Burma. — Dr. A. W. G. Bleeck has 
recently made a careful study of the occurrence of jadeite in the 
Xachin Hills of Upper Burma, and has presented some important 
points in regard to the origin of this much discussed mineral. 
It is found at three places, viz.: Tawmaw, Hweka and Mamon, 
of which the first named is the most important and interesting. 
The jadeite there occurs in a dike of igneous origin which is 
intrusive in serpentine, the predominant rock of the plateau. 
Besides the discussion of the properties of the mineral itself, and 
the species associated with it, the conclusions of the author are 
presented as follows in regard to the question of its origin : "He 
concludes that the jadeite is the result of the metamorphism of 
an albite-nepheline rock originally forming the dike, both 
minerals being found together with the jadeite at Tawmaw. 
The change would be represented chemically as follows : 

NaAlSi0 4 (Nepheline)+NaAlSi q 8 (Albite) = 

'2NaAlSi 2 9 (2 Jadeite). 

"Under certain conditions of crystallization nepheline-albite 
rock might form, while, under conditions of high pressure during 
consolidation or after, jadeite, which has a much lower molecular 
volume, would be produced, the residual molecule forming albite 
or nepheline according to which molecule was in excess in the 
original magma. In the neighborhood of Tawmaw occur various 
crystalline schists which are intruded into by granite. The 
granite is traversed by veins of aplite and pegmatite (products 
probably of the same great eruption) and masses of crystalline 
limestone are found associated with the granite rocks, containing 
various minerals characteristic of contact-metamorphism. The 
relations of the granite to the crystalline limestone in this region 
are similar to those of Mandalay Hill, Sagyin, and Mogok in the 
Ruby Mines district, where similar contact minerals, including 
the different varieties of corundum, are found in the metamor- 
phosed limestone. The crystalline schists include chloritic schists 



34:4 Scientific Intelligence. 

with well-formed crystals of magnetite, actinolite schists and 
glaucophane schists. These are all regarded by Dr. Bleeck as 
the metamorphic products of basic igneous rocks affected by the 
adjoining granite intrusions. The serpentines form a long nar- 
row ridge, flanked on one or both sides by saussuritic-gabbros, 
saussuritic glaucophane-schists and chloritic schists. These rocks 
are traversed by granite and veins of quartz ; all the rocks are 
regarded as genetically related, and as the results of the differen- 
tiation of the same magma, which gave rise successively to the 
peridotites, gabbros, nepheline-albite (jadeite) rock and the 
siliceous end-products of granite and quartz." 

"Dr. Bleeck finds that the eruptives, including the jadeite, are 
prominently represented among the bowlders in the Tertiary 
conglomerate, and thus must have become weathered to con- 
tribute to the Tertiary sediments." — Records Geol. Surv. India, 
xxx vi, pp. 254-285, with five plates ; also xxxvii, 16, 1908. 

11. Rubies from Tipper Burma. — Dr. A. W. G. Bleeck has 
CKamined the ruby deposits of Naniazeik, Myitkyina district, 
Upper Burma, which resemble these of Mogok described by 
J. W. Judd and G. Barrington Brown. The results are slated 
as follows: "Rubies are found in the soil and alluvial accumula- 
tions around the village of Naniazeik as well as in the river- 
gravels on the eastern slopes of the mountain ranges between 
Naniazeik and Manwe. This mountain range is composed mainly 
of granite and crystalline limestone, the latter having obtained 
its crystalline characters probably as stated before, through the 
intrusion of the granite. The limestone contains various 
minerals as the result of contact-metamorphism— garnet, spinel, 
chondrodite, graphite, forsterite, and other accessories, besides 
the valuable rubies and sapphires. The contact of granite and 
marble, exposed on the road from Sikaw to Naniazeik, shows the 
granite to assume a pressure structure near the margin, and to 
contain large quantities of phlogopite, which is a prominent 
mineral also on the marble side of the contact. The marbles, 
when freshly broken, have the characteristic evil smell of many 
limestones charged with nitrogenous organic matter. The marble 
is thus probably the result of the metamorphism of an ordinary 
sedimentary limestone of chetnico-organic origin, but no data are 
obtainable to determine its age." — Records Geol. Surv. India, 
xxxvi, pt. 3, xxxvii, pt. 1, p. 18. 

12. New Group of Manganates. — L. Leigh Feemor, follow- 
ing Laspeyres, concludes, from a study of psilomelane and related 
compounds, that there is a special family of manganates, corre- 
sponding to the acid II 4 Mn0 5 . This includes psilomelane, the 
lead manganate coronadiie of Lindgren and Hillebrand and the 
barium-iron maganate hollandite, from Kajlidongri and else- 
where in Central India. — Records Geol. Surv. India, xxxvi, 295, 
xxxvii, 16. 

13. Die Rlutenpflanzen Afrikas: eine Anleitung zum Restim- 
men der Gattungen der afrikanischen Siphonogamen ; by Franz 
Tiionner, Pp. xv i, 672., with 150 plates and a map of Africa, 



Geology and Natural History. 345 

showing the floral regions according to Engler. Berlin, 1908 (R. 
Friedlander & Sohn). — Although the flora of Africa has received 
a great deal of attention from systematic botanists, especially in 
Europe, there has been lacking a comprehensive work dealing 
with the plants of the entire continent. The present volume sup- 
plies this want, and is intended more particularly for the use of 
travelers and colonists, giving them the means whereby they may 
ascertain the names and relationships of the plants they meet 
with. For this reason it includes not only indigenous plants but 
also many that have been introduced, especially those that have 
become naturalized or are cultivated on a large scale. The work 
first gives a series of keys for the determination of the natural 
families. This is followed by a full characterization of the fami- 
lies, under which further keys are given leading to the determina- 
tion of the genera. The families are arranged according to the 
widely adopted Engler-Prantl system, and under each family the 
number of genera and species- represented in Africa is noted. 
Although many of the generic characters are indicated in the 
keys, the genera themselves are not formally described, and no 
attempt is made to enumerate the peculiarities of the species. It 
is only in the case of economic plants, in fact, that the names of 
the species are given. In the numerous plates at the close of the 
volume a characteristic species from each of the more important 
families is represented. In an interesting summary the author 
estimates the number of phanerogams for the whole world at 
136,000, of which no fewer than 39,000 are accredited to Africa. 

a. w. E. 

14. Schwendeners Vorlesungen uber mechanische Probleme der 
jBotanik, gehalten an der ZTniversitdt Berlin : by Dr. Carl 
Holtermaxn. Pp. vi, 134, with portrait of Schwendener and 90 
text figures. Leipzig, 1909 (W. Engelmann). — Professor Schwen- 
dener's lectures on the mechanical problems of botany give in a 
condensed form the results of his important investigations in a 
field of botanical science which he has made peculiarly his own. 
These lectures have been edited by his former pupil, Professor 
Holtermann, and are now made accessible to a wider circle of 
botanists through the publication of the present work. Among 
the topics treated, the following are perhaps the most important : — 
the mechanical system (skeleton) of plants, the theory of leaf 
arrangement, the upward flow of sap, the stomata, the twining 
of plants. Wherever possible the mechanical properties of the 
structures discussed are expressed by mathematical formulas. 

a. w. E. 

15. Plant Study, with Directions for Laboratory and Field 
Work; by W. H. D. Meiek, Superintendent City Schools, 

Havana, Illinois. Boston, 1909 (Ginn & Company). — A series 
of loose sheets, held together in a binder ; 36 of the sheets give 
directions for laboratory exercises, abundant space being left for 
notes. The exercises are apparently designed to prepare the stu- 
dent for the analysis of plants, the last part of the series includ- 
ing a number of blanks for plant descriptions, spaces being left 
for the specimens themselves. a. w. e. 



316 Scientific Intelligence. 

III. Miscellaneous Scientific Intelligence. 

1. The Carnegie Foundation for the Advancement of 
Teaching. Third Annual Report of the President ■, Henry S. 
Pritchett, and Treasurer, Thomas M. Carnegie. Pp. 211. 
New York City, 1908. — The total capital of the Carnegie 
Foundation at the end of the year closing on September 30, 1908, 
was nearly $11,000,000, a million dollars having been added by 
successive accumulations since the original gift. The income of 
the year amounted to $530,300, of which $287,000 were expended, 
leaving the balance of $243,200 to be added to the capital. Of 
the amount paid for retiring allowances, $240,600, some two-thirds, 
went to professors, officers and widows in accepted institutions, 
and one-third was paid to those in institutions not on the regular 
list. Seven institutions have been added during the past year, 
making the total now sixty-two. A discussion of the cost of 
maintaining this retiring allowance system brings out the fact 
that the sum now paid is 5 per cent, of the active pay of all the 
professors in service in the sixty-two institutions; it is suggested 
that this amount is likely to increase somewhat in the future, but 
even in that case it seems small in view of the importance 
of the results obtained. The most important change intro- 
duced during the past year has been the admission of state 
universities, colleges and technical schools to the Foundation, 
Mr. Carnegie having added a sum of $5,000,000 to the original 
gift. There are some eighty-three institutions included in the 
list given of tax-supported institutions in this country and 
Canada. It is to be noted, however, that in order that an institu- 
tion should get the benefit of the endowment it must be of the 
requisite academic grade and, further, it is necessary that the 
application made should be approved by the Governor and by a 
special vote of the legislature of the state. 

An interesting work undertaken in connection with the Foun- 
dation is the exchange of teachers between Prussia and the 
United States. This plan involves the sending of a number of 
college or high school teachers from this country to Prussia, and 
the coming of a like number of gymnasium teachers to the 
United States. The instruction is intended to be supplementary 
to that ordinarily given, including the informal teaching of the 
language of the country from which the teacher comes, as also of 
its ideals and customs, school regulations, etc. The teacher him- 
self is expected to gain much in experience and breadth of view 
from his life in the foreign country. The amount of money 
involved is small : for example, Prussian teachers receive leave of 
absence from their government, with pay and traveling expenses, 
receiving, also, from the college or high school in America where 
they are stationed, some $200 or $400, according as the length of 
service is a half year or an entire year. A list of seven Prussian 
teachers is given, who have already been assigned to different 
American schools, and of eight American teachers who have been 



Miscellaneous Intelligence. 347 

accredited to Prussian schools. The next assignment of teachers 
to Prussia will be made in June 1909, for the semester beginning 
with October 1. 

In addition to the interesting detailed statements in regard to 
the special work of administering the Carnegie Foundation, Dr. 
Pritchett contributes also a series of chapters on various univer- 
sity and college questions, which are most suggestive, and which 
should be carefully considered by those engaged in such work. 
Some of the topics discussed are the following : Progress toward 
unity in college requirements for admission ; the admission of 
conditioned and of special students ; class room and laboratory 
instruction by teachers ; the support and organization of higher 
education; the standards of professional education in the United 
States. An account is also given of the various denominational 
boards engaged in regulating education. 

Great as is the value of the work accomplished by the Carnegie 
Foundation for our higher educational institutions in granting 
.retiring allowances to teachers, it is obvious that, under Dr. 
Pritchett's wise and enlightened administration, its usefulness is 
extending out into a much broader sphere, and will bring about 
higher standards and more uniformity in ideals and methods. 

2. Carnegie Institution of Washington. — Recent publications 
of the Carnegie Institution are given in the following list (con- 
tinued from vol. xxvi, pp. 519). 

Year Book No. 7, 1908. Pp. vii, 240, with twelve plates. Feb- 
ruary, 1909. — Noticed on p. 26V. 

No. 85. Index of Economic Material in Documents of the 
United States, California, 1849-1904. Prepared for the Depart- 
ment of Economics and Sociology of the Carnegie Institution of 
Washington; by Adelaide R. Hasse. Pp. 316, 4to. 

No. 90. Guide to the Manuscript Materials for the History of 
the United States to 1*783, in the British Museum, in minor 
London Archives and in the Libraries of Oxford and Cambridge ; 
by Charles M. Andrews and Frances G. Davenport. Pp. 
xiv, 499. 

No. 93. The Rotation Period of the Sun, as determined from 
the motions of the Calcium Flocculi ; by George E. Hale and 
Philip Fox. Pp. 54, with 2 plates, 4 figures. — The investiga- 
tion of calcium flocculi has led to the conclusion that the rota- 
tion periods for different latitudes show the existence of an 
equatorial acceleration similar to that previously observed in the 
case of sun-spots, faculse and the reversing layer. This accelera- 
tion, approximately stated, varies uniformly with the latitude. 
No definite conclusions can be drawn as to the relative velocities 
of the different phenomena named. 

No. 96. Condensation of Vapor as induced by Nuclei and 
Ions. Third Report ; by Carl Barus. Pp. vi, 139. With 49 
text figures. 

No. 97. Supplementary Investigations of Infra-red Spectra : 
Part V, Infra-red Reflection Spectra ; Part VI, Infra-red Trans- 



348 Scientific Intelligence. 

mission Spectra ; Part VII, Infra-red Emission Spectra ; by 
William W. Coblentz. Pp. 183, with 107 figures. 

No. 98. The Topography of the Chlorophyll Apparatus in 
Desert Plants; by William Austin Cannon. Pp. 42, 15 figures 
and 5 plates. The Induction, Development, and Heritability of 
Fasciations; by Alice Adelaide Knox. Pp. 20, 1 figure and 
5 plates. 

Nos. 102, 103. Papers from the Tortugas Laboratory, Depart- 
ment of Marine Biology of the Carnegie Institution of Washing- 
ton, Alfred G. Mayer, Director. In two volumes, Vol. I, pp. 
190 ; Vol. II, pp. 325. — These volumes contain nineteen papers by 
H. E. Jordan, W. K. Brooks, A. G. Mayer and others. 

No. 106. The Gases in Rocks; by Rollix Thomas Chamber- 
lin. Pp. 80, with 2 text figures. Noticed on p. 190. 

3. Report of the Superintendent of the Coast and Geodetic 
Survey, O. H. Tittmann, showing the Progress of the Work from 
July 1, 1907, to June 30, 1908. Pp. 169, 4to, with 9 illustrations 
in pocket. Washington, 1908. — This Report gives a summary of 
the work of the Survey during the last year. The most important 
feature of this is the completion of the reconnaissance for the 
extension of the primary triangulation from the 98th meridian in 
Central Texas across Mexico and California, to the triangulation 
of the same class which extends along the Pacific coast across 
California, Oregon and Washington. This reconnaissance ex- 
tends along the arc of the parallel for a distance of about 1200 
miles. The completion of the triangulation along the 98th 
meridian is also an interesting completion of the year's work 
extending across the country from Canada to Mexico ; this arc 
has been also extended into Mexico. It is further stated that 
Canada has begun a geodetic survey, so that the work of the 
International Geodetic Association for the study of the earth and 
other related problems is being steadily pushed forward. Much 
progress has also been made during the year on the boundary line 
between the United States and Canada, and on the Alaska-Can- 
ada boundary. Appendix 3 (pp. 69-165), by R. L. Faris, gives 
the results of the magnetic observations of the year. 

4. Principal Facts of the Earths Magnetism and Methods of 
Determining the true Meridian and the Magnetic Declination. 
Pp. 96 with 28 figures. Washington, 1908. — The edition of the 
paper issued by the U. S. Coast and Geodetic Survey on the 
Declination Tables for 1902 having been exhausted, the publica- 
tion has now been issued in revised form wilh such corrections 
as were necessary. The scope of the work makes it of great 
general interest, since it is, in fact, a full and readable presenta- 
tion of the principles of the magnetism of the earth, giving a 
digest of the subject, with many illustrations, from the earliest 
historic times. It is not surprising that so interesting a summary 
of this important subject should have found many readers. 



New Circulars. 



84 : Eighth Mineral List : A descriptive list of new arrivals, 

rare and showy minerals. 

85 : Minerals for Sale by Weight: Price list of minerals for 

blowpipe and laboratory work. 

86 : Minerals and Rocks for Working Collections : List of 
common minerals and rocks for study specimens ; prices 
from i^ cents up. 

Catalogue 26: Biological Supplies: New illustrated price list 
of material for dissection ; study and display specimens; 
special dissections ; models, etc. Sixth edition. 

Any or all of the above lists will be sent free on request. We are 
constantly acquiring new material and publishing new lists. It pays to 
be on our mailing list. 



Ward's Natural Science Establishment 

76-104 College Ave., Rochester, N. Y. 

Ward's Natural Science Establishment 

A Supply-House for Scientific Material. 

Founded 1862. Incorporated 1890. 

DEPARTMENTS : 

Geology, including Phenomenal and Physiographic. 
Mineralogy, including also Rocks, Meteorites, etc. 
Palaeontology. Archaeology and Ethnology. 

Invertebrates, including Biology, Conchology, etc. 
Zoology, including Osteology and Taxidermy. 
Human Anatomy, including Craniology, Odontology, etc. 
Models, Plaster Casts and Wall-Charts in all departments. 



Circulars in any department free on request ; address 

Ward's Natural Science Establishment, 

76-104 College Ave., Rochester, New York, V S. A. 



CONTENTS. 

Page 

Art. XX. — Permeabilities and the Reluctivities, for very 
Wide Ranges of Excitation, of Normal Specimens of 
Compressed Steel, Bessemer Steel and Norway Iron 
Rods ; by B. O. Peirce 273 

XXI. — Ice Movement and Erosion along the Southwestern 

Adirondack's ; by W. J. Miller 289 

XXIT. — Estimation of Yanadic and Arsenic Acids and of 
Vanadic and Antimonic Acids, in the Presence of One 
Another ; by G. Edgar 299 

XXIII. — Method for the Iodometric Estimation of Silver 
Based upon the Use of Potassium Chromate as a Pre- 
cipitant ; by F. A. Gooch and R. S. Bosworth ._ 302 

XXIV. — Brown Artesian Waters of Costilla County, Colo., 
their relations to Certain Deposits of Natron or Soda, 
and what they teach ; by W. P. He add en 305 

XXV. — Volumetric Determination of Small Amounts of 

Arsenic ; by L. W. Andrews and H. V. Farr 316 

XXVI. — Preliminary Report on the Messina Earthquake of 

December 28, 1908 ; by F. A. Perret . . _ 321 

XXVII. — New Connecting Link in the Genesis of Fulgur ; 

by C. J. Maury 335 

SCIENTIFIC INTELLIGENCE. 

Chemistry and Physics — Method for Calculating the Boiling-points of 
Metals, F. Krafft and Knocke : Some Properties of the Eadium Emana- 
tion, Rutherford, 336. — Method for Preparing Hydrogen Phosphide, 
Matignon and Trannoy : The Theory of Valency, J. N. Friend : Recent 
Advances in Organic Chemistry, A. W. Stewart, 337. — Michelson's Ether 
Research, E. Kohl : Influence of Pressure upon Thermoelectric Force, H. 
Horig : Aluminum Cell as a Condenser, J. de Modzelewski : Changes in 
the Spectra of Gases submitted to the Magnetic Field, A. Dufour : Die 
elektrischen Eigenschaften and die Bedeutung des Selens fur die Elektro- 
technik, C. Ries, 338. — Physics for Secondary Schools, C. F. Adams : 
Elements of Physics, G. A. Hoadley, 339. 

Geology and Natural History— Iowa Geological Survey Report for 1907, S. 
Calvin : Oklahoma Geological Survey, C. N. Gould, L. L. Hutchison 
and G. Nelson, 339. — Glaciation of Uinta and Wasatch Mountains, W. W. 
Atwood : Glacial Waters in Central New York, H. L. Fairchild : Ground 
Waters of Indio Region, California, W. C. Mendenhall, 340.— Die Alpen im 
Eiszeitalter, A. Penck und E. Bruckner : Geological Survey of Western 
Australia, A. G. Maitland : Die Geologischen Grundlagen der Abstam- 
mungslehre, G. Steinmann, 341. — Mineralien-Sammlungen, W. Prendler: 
Jadeite from Upper Burma, A. W. G. Bleeck, 343.— Rubies from Upper 
Burma, A. W. G. Bleeck : New Group of Manganates, L. L. Fermor : 
Die Blutenflanzen Afrikas, F. Thonner, 344. — Schwendeners Vorlesungen 
uber mechanische Probleme der Botanik, Holtermann : Plant Study with 
Directions for Laboratory and Field Work, Meier, 345. 

Miscellaneous Scientific Intelligence — Carnegie Foundation for the Advance- 
ment of Teaching ; Report of the President, H. S. Pritchett, and Treas- 
urer, T. M. Carnegie, 346. — Carnegie Institution of Washington, 347. — 
Report of the Superintendent of the Coast and Geodetic Survey, O. H. 
Tittmann : Principal Facts of the Earth's Magnetism, 348. 



Dr. Cyrus Adler, 

Librarian U. S. Nat. Museum. 

VOL. XXVII. 



MAY, 1909. 



Established by BENJAMIN SILLIMAN in 1818. 



THE 



AMERICAN 

JOURNAL OF SCIENCE. 



Editor: EDWAKD S. DANA. 

ASSOCIATE EDITORS 

Professors GEOKGE L. GOODALE, JOHN TEOWBEIDGE, 
W. G. FAELOW and WM. M. DAVIS, of Cambridge, 

Professors ADDISON E. VEEEILL, HOEACE L. WELLS, 
L. V. PIESSON and H. E. GEEGOEY, of New Haven^ 

Professor GEOEGE F. BAEKEE, of Philadelphia, 
Professor HENEY S. WILLIAMS, of Ithaca, 
Professor JOSEPH S. AMES, of Baltimore, 
Mr. J. S. DILLEE, of Washington. 



FOURTH SERIES 

No. 161— MAY, 1909. 

VOL. XXVII— [WHOLE -NUMBEK, CLXXVIL] 



NEW HAVEN, CONNECTICUT. 
1909. 



THE TUTTLE, MOREHOUSE & TAYLOR CO., PRINTERS, 123 TEMPLE STREET. 




Published monthly. Six dollars per year, in advance. $6.40 to countries in the 
Postal Union ; $6.25 to Canada. Remittances should be made either by money orders, 
registered letters, or bank checks (preferably on New York banks). 



ANNOUNCEMENT. 



We wish to extend our greeting to the readers of the American Journal 
of Science and give them a hearty invitation to call and see us. We have 
just issued a new 12-page circular of Minerals, etc., and a 10-page circular 
of Rough and Cut Gems. These will be sent free on application. Much 
interesting reading will be found in these circulars. There is an illustrated 
description of the famous Cullman Diamond, which weighs 3,024^ carats; 
also 3 illustrations and description of the beautiful Cinnabars from China ; 
a description of the Benitoite lately discovered in-California and other- new 
discoveries. An illustrated note of the largest reconstructed rubies in the 
rough ever manufactured and other rare minerals described. 

The departments treated are as follows : 

Showy Minerals, 3 

Rare Minerals, 

Polished Minerals, 
Ore Collections, 

Gems, rough and cut, 

Antique Cameos and Mosaics, 

Carvings of Ivory, Quartz and Jade, 
Indian Relics, 

Shells, Corals, etc. 

We are constantly publishing circulars upon new finds of minerals and 
new arrivals, which are sent gratis ; it will pay you to be on our mailing 
list. 



NEW ARRIVALS. 

We have secured many additions to our stock, from many countries, 
among which are many rare examples from the Ural Mts. and Siberia, 
Switzerland, Tyrol, Hungary, Germany, England, East India, Australia, 
Japan, California, Colorado, Arizona, Mexico, Michigan, North Carolina, 
Canada, Nova Scotia, Maine, Pennsylvania, New Jersey, Utah, Missouri, 
Texas, Connecticut, Massachusetts, Montana, New York, etc. 

Write us to-day for our new circulars. 



A. H. PETEREIT, 

81—83 Fulton Street, New York City. 



THE 



AMERICAN JOURNAL OF SCIENCE 

[FOURTH SERIES.] 



Art. XXVIII. — Weathering and Erosion as Time Measures /* 
by Frank Leverett. 

The writer presented a paper, under the above title, at the 
Baltimore meeting of the Geological Society of America, 
which set forth the advantages of estimates of weathering and 
erosion in making correlations between the drift sheets of 
Europe and America. Since the correlation of European 
with American drifts is to be presented as a special paper in 
the international glacial magazine, Zeitschrift fur Grletscher- 
kunde, the present paper will aim only to set forth the value 
of weathering and erosion in making correlations, and will be 
restricted mainly to our American deposits. The topographic 
sheets now published embrace each of the drift formations to 
a sufficient extent to enable the student to use them in com- 
paring glacial formations of different age. They will thus 
serve a valuable purpose in the class room. The present 
paper will fulfill its mission if it helps to a proper understand- 
ing of topographic sheets illustrating each of the glacial drifts. 

There are certain situations in which the student is able to 
determine, by erosion alone, the relative ages of the different 
drift sheets. This can be done where the surface of the drift 
was left in a flat or featureless condition so that the present 
reliefs are almost wholly the result of stream action, and it is 
especially valuable where the streams have been working 
under sufficiently high gradient to admit of the cutting of well- 
delined channels. There are places where stream gradients 
are so low that little or no deepening of the drainage lines has 
occurred. In such situations and in situations where the drift 
topography is of a ridged or complicated character, weathering 
naturally is made a leading criterion for judging of relative 

* Published by permission of Director U.S. Geological Survey. 

Am. Jour. Sci.— Fourth Series, Vol. XXVII, No. 161.— May, 1909. 
24 



350 Leverett — Weathering and Erosion as Time Measures. 

ages of the drift sheets under comparison. The rapidity of 
weathering is especially great in fragmental formations, such 
as the glacial and fluvial deposits, since a multitude of faces 
are exposed for attack. It is thus possible to distinguish 
lesser differences in age than would be detectible in the solid 
formations. Weathering of glacial material or of fluvial mate- 
rial will reach depths to be measured in feet and meters, while 
that in rock formations may reach only a few inches or centi- 
meters. 

There are certain complications in these studies resulting from 
differences in climatic conditions and differences in the consti- 
tution of the drift which have to be properly considered in 
forming the estimates. Leaching goes on so much more 
rapidly under humid conditions than under semi-arid that one 
may easily be misled if he depends entirely upon the depth to 
which the leaching extends. In the same way erosion studies 
are liable to be misleading if one does not take into account 
differences of rainfall, in stream gradient, and in the texture 
of the deposits. It is necessary also, in glaciated regions, to 
ascertain the work accomplished by the drainage from the ice 
sheet, in order to properly compare the work done by the 
present streams on drift sheets of different age. 

In the glaciated portions of North America there are, fortu- 
nately, extensive plains of Kansan, Illinoian, and Wisconsin 
drift and of glacial lake beds, whose conditions of rainfall, 
stream gradient and texture of deposits, are sufficiently alike to 
afford excellent subjects for time estimates based on weathering 
and erosion. It is not claimed that these studies will give time 
measures definitely expressible in thousands of years, at least 
not in the present state of our knowledge ; but it is thought 
that a sufficiently reliable measure of relative age is afforded to 
form a basis for wide correlations, such as those between the 
American and European drift sheets. 

The table here presented gives the names applied to the 
several members of the glacial series in each prominent field 
of Pleistocene glaciation. Some of the lesser fields, west of 
the Rocky Mountains, may have a similar complexity, but this 
has not been so fully established. The Skandinavian field 
has had a great variety of interpretations and stills awaits 
a satisfactory nomenclature. It will, perhaps, serve our pur- 
pose best to make use of the terms employed by the German 
geologists, though here it must be understood that the earlier 
literature is not fully in harmony with the later. The order 
of arrangement in the table is from younger to older in each 
field. 



Leverett — Weathering and Erosion as Time Measures. 351 
Table of drift sheets. 



Keewatiu 


Labrador 


Skandinavian 


Alpine 


Wisconsin 


Wisconsin 


Upper Diluvium 


Wiirm 


Iowan ? 








Illinoian ? 


Illinoian 


Middle Diluvium 


Riss 


Kansan 


Kansan ? 


Lower Diluvium 


Mindel 


Pre-Kansan* 


Jerseyan 




Gunz 




Wisconsin Drift. 





In the first attempts by Chamberlin and others to show the 
complexity of the glacial deposits but two drift sheets were 
recognized, the earlier and the later, and of these the later 
embraces what is now known as Wisconsin drift. Its aspect 
is everywhere much fresher or younger than that of the outly- 
ing and underlying drift sheets, so that it can easily be distin- 
guished from them. With the progress of the study it became 
evident that considerable complexity of ice movement was 
involved in producing the many moraines of this later drift. 
The moraines do not form throughout a concentric series, such 
as one would expect to find had the ice made a simple retreat 
from its field, w T ith halts at places where moraines occur. 
Instead, there are groups of concentric moraines which have a 
trend out of harmony with other groups, resulting in a partial 
overriding of the earlier by the later. From this it is inferred 
that the ice made shiftings in its directions of movement, 
such as call perhaps for some change in its center of accumu- 
lation. 

The earliest group of these moraines of later drift has been 
termed the Earlier Wisconsin, while the remainder of the 
moraines are thrown together under the name Later Wiscon- 
sin. The principal exposure of the Earlier Wisconsin is in 
Illinois, but it has a slight extent beyond the Later Wisconsin 
in Indiana and Ohio and possibly also in Wisconsin. Whether 
it is exposed in states farther w T est has not been determined. 
The amount of weathering and erosion which the Earlier 
Wisconsin has undergone differs so little from that experi- 
enced by the Later Wisconsin that we may feel sure that 
no interval of great length separates them. The knolls 
and basins along the moraines of the Earlier Wisconsin are 
toned down by erosion and filling so that only the larger ones 
are well-defined, whereas in the Later Wisconsin the small 
hummocks and basins still preserve their sharpness of contour. 
But in the Earlier Wisconsin as in the Later there are wide 
areas not yet invaded by drainage lines, as may be seen by 
reference to the Danville, the Urbana, and the Mahomet, 

* The Pre-Kansan is sometimes termed Sub-Aftonian. 



352 Leverett— Weathering and Erosion as Time Measures. 

Illinois, topographic maps. If one should place these maps 
besides those of quadrangles inside the Later Wisconsin, as for 
example the Marion, Sycamore, Upper Sandusky and Arling- 
ton sheets in northwestern Ohio, it would be difficult to dis- 
cover any difference in the character of drainage. The amount 
of weathering, however, is perceptibly greater in the Earlier 

Fig. 1. 




Fig. 1. Part of the Wayne, Michigan, topographic sheet, illustrating the 
slight amount of erosion characterizing the beds of the great glacial lakes 
Scale, 1 : 90000. Contour interval, 20 feet. 

than in the. Later Wisconsin, it being rare to find unleached 
till on the plains of the Earlier Wisconsin at a depth of less 
than a meter, whereas on the Later Wisconsin it is frequently 
found at \ meter or less, and the average depth of leaching in 
the Later Wisconsin scarcely reaches a meter. It is, therefore, 
chiefly by the weathering and by toning down of the slight 
inequalities that the Earlier Wisconsin is distinguished from 
the Later. 



Lever ett — Weathering and Erosion as Time Measures. 353 

The erosion of the Later Wisconsin drift, compared with 
that of the beds of the large glacial lakes which occupied 
its surface during the recession of the ice from the St. Lawrence 
basin, is sufficiently more advanced to be perceptible on many 
of the topographic maps. The sections of the two maps here 
introduced serve to bring out this difference (see figs. 1 

Fig. 2. 



Fig. 2. Part of London, Ohio, topographic sheet, illustrating the amount 
of erosion ordinarily found on a plain of Wisconsin drift where the slope 
is sufficient to favor a somewhat rapid development of drainage lines. The 
incipient stage of development testifies to the recency of the Wisconsin 
glaciation. Scale, 1 : 93700. Contour interval, 20 feet. 

and 2). Fig. 1, which represents a portion of the lake plain 
in southeastern Michigan, shows the drainage lines to have 
developed very few tributaries as well as to have cut only 
insignificant valleys. Fig. 2, which represents the Wisconsin 
drift in the western portion of Scioto basin in Ohio, shows a 
slightly deeper valley cutting and the beginning of numerous 



35-i Lever ett — Weathering and Erosion as Time Measures. 

tributary valleys. The tributaries, however, are in a very 
incipient state of development. This map near West Jefferson, 
Ohio, serves to indicate that fully nine-tenths of the drift plain 
still stands at its original level, and this is true over a large 
part of the surface of the Wisconsin drift. In fact there are 
considerable areas in which less than one-tenth of the surface 
has been lowered or reduced in any way by drainage lines. 

Areas in which the topography of the Wisconsin drift and the 
Post-Wisconsin erosion may be studied without much compli- 
cation with rock topography have already been mapped in 
nearly every state from Massachusetts to North Dakota. 
Thus a considerable part of southeastern Massachusetts, includ- 
ing the islands of Marthas Yineyard and Nantucket, much of 
Long Island in New York, as well as the greater part of the 
plain east and south of Lake Ontario, nearly all the northwest 
one-fourth of Ohio, the quadrangles mapped in southeastern 
Michigan, northeastern Illinois and southeastern Wisconsin 
(except the Broadhead, Janesville and Shoppiere sheets in 
Wisconsin), the area around Minneapolis and St. Paul, Minne- 
sota, the area north and west of Des Moines, Iowa, certain 
sheets in the James Biver valley in South Dakota, and those 
in the eastern part of North Dakota, may be drawn upon for 
illustrations. It will be observed that the morainic areas of 
southeastern Michigan, southeastern Wisconsin, and in the 
vicinity of St. Paul, Minnesota, embrace much poorly drained 
land represented as swamp, whereas such swamp areas are of 
small extent in northwestern Ohio, northeastern Illinois, and 
in the region bordering Lake Ontario. This should not be 
interpreted to indicate that the less swampy areas owe their 
condition to greater age. It results instead from a more favor- 
able original condition for shedding the water or from a differ- 
ent ground-water condition. 

There are extensive areas in New England, as also in New 
York and in Ohio, where rock ridges are prominent, in which 
the Wisconsin glaciation has served to produce what has been 
termed by Salisbury superimposed youth.* In these areas the 
drainage lines are often as poorly developed as in areas of the 
Wisconsin drift, where the rock irregularities are completely 
buried, and they indicate to the physiographer a similar youth- 
fulness. Such areas, however, are not so satisfactory as the 
smooth areas of heavy drift for determining the relative age of 
the different drift sheets. 

The character of the glacial drainage of the Earlier Wis- 
consin appears to have been different from that of the Later 
Wisconsin, there being no extensive outwash gravel plains 
associated with the Earlier Wisconsin moraines such as charac- 

* Journal of Geology, vol. xii, pp. 711-713, 1904. 



Leverett — Weathering and Erosion as Time Measures. 355 

terize portions of the Later Wisconsin moraines of the same 
general region. This fact, and the fact that the outer portion 
of the Earlier Wisconsin drift in Illinois carries a thin coating 
of loess, is thought to indicate that a certain degree of aridity 
prevailed during the culmination and the withdrawal of the 
ice of this early part of the Wisconsin stage of glaciation. 
The relatively arid conditions may account to some degree for 
the poor development of drainage lines, which otherwise seem 
rather inconsistent with the amount of toning down of knolls 
and filling of basins which took place on the Earlier Wisconsin 
moraines. The filling of basins would be accomplished rap- 
idly under conditions of aridity, as is exemplified in our arid 
western region. It is well, therefore, to keep this climatic 
factor in mind when making a comparison of the amount of 
erosion of the Earlier Wisconsin drift with the somewhat 
younger Later Wisconsin drift. 

As may be seen from the above table (p. 351), the Wisconsin 
drift is the only one of the series named in the Labrador and 
Keewatin fields which is correlated with certainty. There is 
a continuous belt of moraines connecting the Labrador and 
Keewatin fields, but these moraines are of Later Wisconsin 
age, and it yet remains to be determined whether the Earlier 
Wisconsin drift is exposed outside the Later Wisconsin in the 
Keewatin field. 

Probable European Correlatives of the Wisconsin Drift. 

In the European fields, Skandinavian and Alpine, and also 
in Great Britain, there is a drift sheet which corresponds 
closely in the amount of weathering with our Later Wisconsin 
drift, while certain of the outer moraines of the north Ger- 
man lowland and of Great Britain seem to correspond with 
our Earlier Wisconsin. The complexity of the last stage of 
glaciation seems, therefore, about as great in northwestern 
Europe as in North America. In the Alps the Wiirm drift 
has a freshness of contour very similar to that of the Later 
Wisconsin moraines, while the " Verwaschene Wiirm," which 
lies outside the Wiirm proper at a few points in northern 
Switzerland, and also at a few points on the southern side of the 
Alps, has a similar toning down to that displayed by our 
Earlier Wisconsin moraines, of which it may prove to be the 
correlative. 

The Upper Diluvium of the JSforth German Lowland is much 
more extensive than the limits assigned by Geikie to the fourth 
or Mecklenburgian drift. The great Baltic ridge seems, as 
has been interpreted by several German geologists, to lie well 
within the limits of the last glaciation, though it certainly 



356 Leverett — Weathering and JLrosion as Time Measures. 

marks a very important position of the ice border, whether a 
mere halt or a readvance is not yet clear. The results of my 
own studies are in full accord with those of the German stu- 
dents in placing the limits of the last glaciation far enough 
south to embrace the prominent moraines along the south 
border of the region characterized by lakes. 

Illinoian Drift. 

The Illinoian drift of the Labrador icefield is exposed out- 
side the limits of the Wisconsin drift in Illinois, Indiana, and 
Ohio, and at its farthest extent reached slightly into Iowa.* 
The entire outlying portion of this drift is covered by loess 
except a small district in northwestern Illinois and southern 
Wisconsin, where the loess is of rather patchy distribution. 
Beneath the loess there are widespread occurrences of a black 
soil and muck known as the Sangamon soil, which rests upon 
the leached surface of the Illinoian drift. It thus appears 
that the Post-Illinoian history is somewhat complex. The 
occurrence of the black soil and muck is suggestive of cool, 
damp conditions, while the overlying loess suggests conditions 
of aridity. The loess is found to pass beneath the outermost 
moraines of the Earlier Wisconsin drift and hence antedates 
the Wisconsin ice invasion.f 

The leaching which the Illinoian drift has experienced 
appears to have been largely accomplished prior to the deposi- 
tion of the loess, for the loess is sufficiently thick over much 
of the Illinoian area to preserve an unleached basal portion. 
Yet in spite of this cessation of the leaching at the time of 
the loess deposition it had reached a depth, and weathering 
bad gone to a degree, far in advance of the leaching and 
weathering thus far experienced by the Wisconsin drift. The 
removal of calcareous constituents, pebbles as well as fine 
material, was nearly complete to a depth of 4 to 6 feet, while 
weathering often extended to two or three times these depths. 
Indeed pipes and weathered seams extend far down into the 
unleached portion, as is not found to be the case in the Wis- 
consin drift. One cannot well escape the conclusion that the 
time involved in this leaching was at least two or three times 
as great as that of the Wisconsin, and it may considerably 
exceed this amount. In addition to the weathering and leach- 
ing there is a general induration of the entire deposit, pro- 
duced probably by a partial cementation, a feature which is 
very sparingly exhibited by the Wisconsin drift. 

Turning now to the amount of erosion which the Illinoian 

* See The Illinois Glacial Lobe, Monograph xxxviii, U. S. Geological 
Survey, 1899, pp. 24-130. Also Monograph xli, 1902, pp. 253-294. 
f See Monograph xxxviii, pp. 128-129 and plate xi. 



Leverett — Weathering and Erosion as Time Measures. 357 

drift displays, we find that where drainage gradients are good 
it far exceeds that displayed by the Wisconsin drift. An 
excellent topographic sheet for comparing the erosion of the 
Illinoian with the Earlier Wisconsin is that of the Peoria, 
Illinois, quadrangle, the western portion of which shows Post- 
lllinoian erosion, while the eastern portion shows Post- Wiscon- 
sin erosion, in studying this sheet, however, it should be 
borne in mind that the Wisconsin drift area is morainic and 
more elevated than the Illinoian, and thus has a great advan- 
tage in the rapidity of shedding its water ; but in spite of this 
advantage the extent of the erosion is far less than in the 
Illinoian drift. 

Topographic sheets that lie entirely within the Illinoian 
drift area are the Springfield, Eldorado and Belleville, Illinois, 
sheets. The Belleville topographic sheet is of especial inter- 
est since it exhibits the contrast in the work of streams under 
different gradients. The western portion, a section of which 
is here reproduced in fig. 3, illustrates the appearance of the 
drainage when under a good gradient, while fig. 4, taken from 
the southeast part of the same quadrangle, illustrates drainage 
under a low gradient. It will be observed that Silver Creek 
(in fig. 4) has a broad valley in which abandoned courses of 
the stream testify to the sluggishness of the drainage and sug- 
gest that aggradation rather than channeling is now in progress. 
The entire Eldorado topographic sheet is in a region of low 
drainage gradient, while the Springfield sheet represents a 
region of more rapid drainage gradient. The series is, there- 
fore, very instructive in setting forth the amount of variation 
that may be found on a drift formation of a given age. 

To properly compare the Illinoian erosion with the Wiscon- 
sin one must select areas in which drainage conditions are 
similar, and also where no inequalities of drift surface, such as 
moraines, etc., come in to give complexity. Figs. 2 and 3 
represent fairly well the development of drainage under sim- 
ilar conditions of slope. The conditions are perhaps more 
favorable in the area represented by fig. 2 for the rapid devel- 
opment of drainage than in fig. 3, yet the stage of drainage 
development in fig. 3 is advanced far beyond that in fi.g. 2. 
The writer has estimated that in the portion of the Illinoian 
drift in western Illinois and southeastern Iowa, where the con- 
ditions for development of drainage are the best, that approxi- 
mately one-half the surface has been reduced below the origi- 
nal level as a result of the drainage, while in the Wisaonsin 
drift, as already indicated, scarcely one-tenth of the surface has 
been thus reduced. It is evident that the time required for 
developing the drainage found on the Illinoian is much greater 
than that on the Wisconsin on the assumption that the climatic 



358 Lever ett — Weathering and Erosion as Time Measures. 

conditions have been uniform and similar to the present con- 
ditions. It is probable, however, that a period of aridity came 
in at the time of the loess deposition, and this would lengthen 
materially the time required for the development of the drain 
age lines in the Illinoian drift. 

The presence of a drift sheet of Illinoian age in the Kee- 
watin field is still open to question. The only deposits sus- 

Fig. 3. 




Fig. 3. Part of the Belleville, Illinois, quadrangle, showing Post-Illinoian 
drainage under a fair stream gradient. Scale, 1 : 93700. Contour interval, 
20 feet. 

pected to be of Illinoian age, so far as yet noted, are found in 
southeastern Minnesota and neighboring parts of Wisconsin, 
just outside the limits of the Wisconsin drift, and these have 
not as yet been given sufficiently close study to clear up their 
relations to the broad sheet of Illinoian drift of the Labrador 
field. 



Leverett — ■ Weathering and Erosion as Time Measures. 359 



European Correlatives f of the Illinoian Drift. 

From their position in the series one would naturally expect 
to find correlatives of the Illinoian drift in the Middle Dilu- 
vium of north German lowland and the Riss drift of the 
Alpine region. But neither the Riss nor the Middle Diluvium 
was found to bear a close resemblance to our Illinoian drift. 
These European drifts are commonly weathered to a depth of 

Fig. 4. 




Fig. 4. Part of the Belleville, Illinois, quadrangle, showing Post-Illinoian 
drainage under a very low stream gradient. Compare the small amount of 
dissection of the drift plain with that in fig. 3. Scale, 1 : 93700. Contour 
interval, 20 feet. 

only 3 or 4 feet and there is a very abrupt change to fresh- 
looting drift at the base of the weathered portion. Were the 
upper part to the depth of a meter stripped off from these 
drifts one would find difficulty in distinguishing the Kiss from 
the Wiirm, or the Middle Diluvium from the Upper Diluvium. 
The induration of the deeper portion, so generally found in 



360 Leverett — Weathering and Erosion as Time Measures. 

the Illinoian, is not seen in the Riss and the Middle Diluvium, 
nor is there the deep penetration of weathered pipes and seams 
into the unweathered portion. The Kiss moraines have not 
the sharp hummocks that characterize the Wiirm moraines, but 
the contrast is not much more striking than is found between 
our Earlier and Later Wisconsin moraines. The impression one 
obtains is that the Riss and the Middle Diluvium have reached 
a condition of weathering intermediate between our Earlier 
Wisconsin and our Illinoian drifts. Further study into cli- 
matic conditions in the several fields and into the constitution 
of the drift deposits may serve to remove some of the difficulties 
of correlation. Portions of the area of the Middle Diluvium 
lying east of the Harz Mountains have at present a very light 
rainfall, which seems to produce less leaching than is accom- 
plished in more humid districts to the west. Possibly long 
continued aridity will account for the lack of deep penetration 
of weathered pipes and seams. At present waters charged 
with calcareous constituents rise into the soil by capillary 
action and tend to keep it well supplied with calcareous mate- 
rial. Prof. Keilhack reports that the calcareous material is in 
the form of a coating on the grains of silica, feldspar, etc., and 
not in the form of grains or coarse particles of limestone. In 
other words, the present calcareous content is secondary. 

So far as inferences can be drawn from relative amounts of 
erosion these European drifts differ but little from the Illi- 
noian in the places where erosion conditions are similar as to 
gradient and texture of deposit. The erosion of the high ter- 
races (Hochterrassen), which connect with the Riss moraines 
of the Alpine region, seems to be about as great as the ordi- 
nary erosion of the Illinoian drift under good stream gradients. 
The erosion of the Middle Diluvium has been sufficient to 
remove nearly all traces of the deposit along the borders of 
the main drainage lines, which compares favorably with the 
amount of erosion in our Illinoian drift. 

Kansan Drift. 

The Kansan drift, which is the most extensive sheet of 
the Keewatin field, is widely exposed in northern Missouri, 
southern and western Iowa, and neighboring parts of Kansas 
and Nebraska. Its northern portion is concealed by the Wis- 
consin drift. The greater part of this drift, like the Illinoian, 
is covered by loess. Some of this deposit may be older than 
the Illinoian drift, at least loess has been found between the 
Kansan and the Illinoian in the district along the Mississippi 
where the Illinoian overlaps the Kansan drift. A large part 
of the loess, however, seems to be of the same age as that which 



Leverett — Weathering and Erosion as Time Measures. 361 

covers the Illinoian drift, and seems to have been deposited 
after the Kansan drift had been very greatly eroded, for it 
covers the valley terraces and valley slopes as well as the 
uplands, and appears deep down in the valleys in positions that 
can scarcely have been reached by a secondary deposition or 
by a creep of the deposit. Beneath the loess on the remnants 
of the old drift plain, preserved between the valleys, one finds 
a black soil and muck similar to that at the junction of the 
loess and the Illinoian drift. The region seems, therefore, to 
have had alternations of wet and dry climates which will compli- 
cate to some extent the estimates of erosion. So far as the 
time element is concerned the periods of aridity will render 
the length of the Post-Kansan time greater than that required 
for the mere cutting of the valleys under present conditions of 
precipitation. The presence of two sheets of loess would seem 
to call for two periods of aridity. 

The topographic maps of northeastern Missouri, from which 
figs. 5 and 6 are taken, represent a region where the erosion 
is almost entirely in glacial deposits, the rock surface being 
deeply buried. It will be observed that a large part of the 
original Kansan drift plain has been destroyed, the plain being 
preserved only in narrow tabular divides, such as that on 
which the towns of Atlanta and Memphis stand. The valley 
slopes as well as valley bottoms are very broad. The slopes 
are so toned down as to be generally below an angle of 5° and 
not uncommonly are reduced to 3° or even less. By reference 
to figs. 5 and 6 it will be seen that the 20-foot contours are 
generally broadly spaced, so that four contours, or 80 feet 
elevation, occupy a breadth of one-fourth to one-half mile on 
a valley slope. On streams of comparatively small drainage 
area the valley bottoms frequently exceed a mile in breadth. 
In the region covered by these maps and in much of northern 
Missouri and southern Iowa narrow tabular remnants of the 
original drift surface are preserved. But as one passes from 
the tributaries of the Mississippi westward to the tributaries 
of the Missouri these tabular remnants gradually dwindle and 
in the vicinity of the Missouri valley entirely disappear. The 
tributaries of the Missouri have some advantage in directness 
of drainage, which probably partially accounts for the more 
advanced state of drainage erosion. There may also have been 
a flatter surface in the vicinity of the Mississippi, since the 
underlying limestone there was left with broad tabular surfaces 
at the oncoming of glaciation, while the sandstone areas to the 
west were broken up into ridges and hills. The tabular divides 
in the drift surface, however, do not coincide with preglacial 
divides in the limestone surface, nor are they restricted to the 
limestone area, but extend westward some distance into the 



362 Lever ett — AYeatliering and Erosion as Time Measures. 

sandstone area. The average amount of removal of drift in 
Post-Kansan time is estimated to be somewhat more than 
50 feet, at least it would take more than that amount to bring 
the valleys up to a level with the tabular divides. This is 
apparent from the sections of maps here presented, and was 
carefully worked out in the field in a neighboring county of 
Iowa by Dr. C. H. Gordon, while connected with the Iowa 

Fig. 5. 



Fig. 5. Part of the Atlanta, Missouri, topographic sheet showing Post- 
Kansan drainage. Scale, 1 : 86500. Contour interval, 20 feet. 

Geological Survey. To the question whether the streams 
may have begun their work in depressions somewhat below the 
level of the tabular divides, it is remarked that the lack of 
correspondence between the tabular drift divides and the 
preglacial rock divides, together with the consistent fitting in 
of each tabular divide with its neighbors to form a plain slop- 
ing in the direction of drainage, as well as the evidence to be 
gathered from the lateral drainage, gives a strong presump- 
tion that the surface was brought up very nearly to a plain of 
the kind preserved in the tabular remnants. 



Leverett — Weathering and Erosion as Time Measures. 363 

The Kansan drift, in places where erosion has been at a 
minimum, as on the tabular divides, has a weathering to a 
depth of several meters. There is more or less complete 
removal of limestone pebbles as well as of finer calcareous 
material to an average depth of 1-J to 2 meters. Apparently 
the deposit originally was as calcareous at its surface as in its 
unleached portion. It lies in a region where limestone forma- 
tions have extensive outcrops and there are numerous calcareous 

Fig. 6. 




Fig. 6. Part of Edina, Missouri, topographic sheet showing Post-Kansan 
erosion. Scale, 1 : 173000. Contour interval, 20 feet. The map on this 
scale is introduced for purposes af comparison with that of the so-called 
Iowan, fig. 7, which has the same scale. 

nodules at the base of the leached part testifying to a contribu- 
tion of lime from it. The uppermost half meter of the leached 
portion is often of a decidedly red tinge, below which is a 
transition through brown to yellow and finally to gray or blue 
in a space of 4 to 6 meters or more. The statements just made 
represent conditions in the humid Mississippi valley where 
there is an average rainfall of about 40 inches. But as one 
passes westward into the semi-arid districts along and west of 
the Missouri River, the amount of leaching in this drift sheet 
is found to become markedly less, and that too where erosion 
has been fully as great as in the Mississippi valley. 



36tt Lever ett — Weathering and Erosion as Time Measures. 

The best sheets to illustrate the Post-Kansan erosion are the 
Atlanta, Edina, and Kahoka, Missouri, quadrangles, not only 
because the drift there is sufficiently thick to dominate the 
topography but also because the maps are better executed than 
older contour maps of eastern Iowa and eastern Nebraska. 
The DesMoines, Iowa, and Waukee, Iowa, topographic sheets 
are, however, recent and of a high standard, and they serve to 
exhibit the contrast between the erosion of the Kansan and 
the "Wisconsin, the southern edges of the maps being within 
the limits of the Kansan drift, while the remaining, much less 
eroded portion is in the Wisconsin drift. 

As to the length of the Post-Kansan period of weathering 
and erosion compared with that of the Post- Wisconsin, only a 
rough approximation is likely to be reached by measurements 
of the changes effected in the two drifts. Dr. H. F. Bain 
has made a computation of the relative size of valleys in the 
Kansan and in the Wisconsin drift in the vicinity of Des 
Moines, Iowa, and ascertained the capacity of the Post-Kansan 
valleys to be about seventeen times that of the Post- Wisconsin.* 
While a considerable time must necessarily be allowed for a 
drainage system to get well started and thus correspondingly 
lengthen the Post-Wisconsin time and reduce the difference in 
the age of the Wisconsin and the Kansan drift, the periods of 
aridity through which the Kansan drift has passed prior to the 
Wisconsin stage of glaciation may fully offset this lengthening 
of the Post-Wisconsin time. 

The question of a correlative of the Kansan drift in the 
Labrador field is considered below in connection with the dis- 
cussion of the Pre-Kansan drift. 

European Equivalents of the Kansas Drift. 

. In each of the glaciated districts of Europe a drift sheet is 
found in which the weathering is strikingly similar to that of 
the Kansan drift. In the Alpine region the apparent correla- 
tive is the Mindel drift ; in north Germany it is the Lower 
Diluvium, the drift formed at the maximum extent of the 
Skandinavian ice held, when it reached the recesses of the 
mountainous districts of Silesia and Saxony ; in Great Britain 
it is the chalky bowlder clay of the southeastern counties. Not 
only is there a similar depth of leaching but also a similar 
degree of oxidation, and a deep penetration of weathered pipes 
and seams. In the Cromer sections on the east coast of Eng- 
land the weathered seams are conspicuous to depths of 10 to 15 
meters, and the leaching of the chalky bowlder clay is generally 
thorough to a depth of about 2 meters. The reddish tinge of 
the upper part of the weathered drift in these supposed Euro- 

* Geology of Polk County, Iowa Geological Survey, vol. vi. 



Leverett — Weathering and Erosion as Time Measures. 365 

pean correlatives is also as conspicuous as in our Kansan. If 
we consider that since the time when the Kansan drift and its 
correlatives were deposited there have been repeated climatic 
changes from cold to warm and probably also from humid to 
arid, it may seem remarkable that weathering should have been 
so similar in amount in these widely separated glaciated dis- 
tricts. It is at least suggestive of worldwide influences in the 
control of the climatic conditions that produced glaciation. 

Pre-Kansan Drift. 

The Pre-Kansan drift of the Keewatin field, so far as yet 
known, lies entirely within the limits of the Kansan drift and 
has exposures only beneath the Kansan. It is from these 
exposures alone that one may learn the degree of weathering, 
and they are inadequate to throw much light upon its erosion. 
Recent studies in western Iowa, under the Iowa Geological 
Survey, indicate that herbivorous animals had taken possession 
of that region in the interval of ice recession between the Pre- 
Kansan and the Kansan, so it is inferred that there was plant 
growth such as comes only with a genial climate. Data bear- 
ing directly upon the amount of leaching of the gravelly sur- 
face portion of this drift were obtained by my assistant, P. T. 
Chamberlin, in 1907, at exposures near Afton Junction, Iowa, 
the type locality of the interglacial Aftonian exposures. Four 
classifications of pebbles from near the surface of the deposit 
show percentages of limestone pebbles ranging from 14 to 20 
per cent, while two classifications of pebbles from underlying 
till, not affected by leaching, show 38 and 52 per cent of 
limestone pebbles. In other respects the pebbles of the gravel 
and o*f the underlying till are strikingly similar. It thus 
appears probable that the limestone content of the gravel was 
originally about the same as that of the till, and that nearly 
two-thirds of the limestone had been removed from the surface 
portion of the Pre-Kansan drift before it was buried under the 
Kansan. The limestone pebbles remaining in this surface por- 
tion are in an etched and deeply weathered condition, so that 
had there been but a moderate extension of the interglacial 
stage the limestone might have been completely removed. In 
places gravelly portions of the Pre-Kansan drift had become 
firmly cemented in this interglacial stage, so that masses were 
gathered up by the ice of the Kansan stage and incorporated 
as bowlders in the Kansan till. The change which the Pre- 
Kansan drift experienced probably approaches if it does not 
equal that which the Illinoian has experienced, and appears to 
be greater than that which has affected the Wisconsin drift. 
It may therefore properly be referred to a distinct glacial stage. 

Am. Jour. Sci.— Fourth Series, Vol. XXVII, No. 161.— May, 1909. 
25 



366 Leverett — Weathering and Erosion as Time Measures. 

Equivalents of Kansan or Pre-Kansan in the Labrador Field. 

The writer lias never had opportunity to study, or even to 
travel across, the area in New jersey and northeastern Penn- 
sylvania occupied by the sheet of old drift of the Labrador 
field to which Chamberlin and Salisbury have applied the term 
Jersey an. The patchy distribution and weathered appearance 

Fig. 7. 




Fig. 7. Part of Oelwein, Iowa, topographic sheet showing topography of a 
so-called Iowan drift plain. The amount of erosion here displayed should 
be compared with that in the Illinoian, figs. 3 and 4, and Kansan drift, iigs. 
5 and 6. Scale, 1 : 173000. Contour interval, 20 feet. 

reported by those who have studied it, seem to indicate an age 
as great as that of any drift sheet in the Keewatin field. It is, 
therefore, considered by these students a probable correlative 
of the Pre-Kansan or Sub-Aftonian drift of the Keewatin field. 
The old drift of northwestern Pennsylvania, which has been 
examined by the present writer, is thought from the weather- 



Leverett — Weathering and Erosion as Time Measures. 367 

ing and erosion to be at least as old as the Kansan drift, and 
possibly the correlative of the Pre-Kansan drift. Its aspect 
is decidedly older than the Illinoian drift exposed in the west- 
ern part of the field covered by the Labrador ice sheet. The 
topography of the quadrangles within the limits of the Jer- 
seyan drift and of the old drift of northwestern Pennsylvania 
is controlled to so slight an extent by glacial deposits that the 
topographic sheets are not likely to have much value in setting 
forth the amount of Post- Jersey an erosion, and hence are not 
drawn upon for illustrations. 

European Equivalents of the Pre-Kansan. 

Upon turning to the European fields conditions somewhat 
similar to that of the Kansan and Pre-Kansan have been 
brought to notice. Thus in the Alpine region the oldest drift, 
known as the Gunz, is almost completely overridden by another 
old drift, the Mindel. In that region, as in the American, 
just discussed, it is difficult to determine the difference in age 
because of the meager exposure of the Gunz drift. The 
studies of Penck and Bruckner, however, have shown that 
there was an advanced state of weathering of the surface of 
the Gunz drift before the Mindel glaciation buried it. The 
Mindel and Gunz drifts are thought by Penck to be separated 
by a shorter interval than that between the Piss and Mindel.* 
In the Skandinavian field observations by Nathorst and others 
indicate that in southern Sweden a set of .striae out of harmony 
with the ice movement which carried the lower bowlder clay of 
Germany is likely to have preceded that movement. While 
the earlier movement may not have been sufficiently distinct 
from the later to constitute a separate stage of glaciation, 
Geikie has suggested applying to it a distinct name, Scanian, 
from the district in Sweden where it was brought to light. 

Iowan drift? 

Since it is through a study of weathering and erosion that 
the presence of the so-called Iowan drift has come to be 
questioned, it seems fitting that this discussion should include 
a brief statement concerning the leading features in the region 
that drift has been supposed to cover. The region lies in 
northeastern Iowa immediately east of the area covered by the 
Des Moines lobe of Wisconsin drift. The topographic maps 
which lie wholly or partly within its limits are the Decorah, 
Oelwein, Elkader, Winthrop, Farley, Stan wood and Rock 
Island. Pig. 7, which represents part of the Oelwein quad- 

* Die Alpen im Eiszeitalter, p. 1168. 



368 Leverett — Weathering and Erosion as Tim,e Measures. 

rangle, illustrates well the topography of the part of this 
region where the valleys are excavated entirely in the drift. 
Farther east the drift becomes so thin that rock ridges dominate 
the topography. It will be observed , that the surface presents 
broad slopes, as in the Kansan drift, but the valley bottoms 
are less well defined, the slopes on either side extending nearly 
to the streams. Tabular divides are also wanting and con- 
siderable inequality is found along the lines of water parting. 
This region differs also from neighboring parts of the Kansan 
in being nearly free from loess. 

The inspection of such a map as that given in fig. 7 naturally 
raises the question whether the departures which it exhibits 
from the typical Post-Kansan erosion, shown in figs. 5 and 6, 
may be interpreted as a superimposed youth, such as might 
result from a fresh invasion of the ice into a much eroded 
region of older drift. A study in the field would naturally 
be directed to the solution of the question of the existence of 
a drift sheet too fresh to be consistent with the interpretation 
that the valleys were entirely formed after its deposition, and 
a special search should be made to discover whether a fresh 
drift is so situated in the valleys as to indicate clearly that it 
was deposited after the valleys were excavated. Upon visit- 
ing the field in 1906 and again in 1907, and spending about 
two months in careful examination of the exposures on uplands, 
slopes, and valley bottoms, the writer was unable to discover 
any trace of fresh drift. The valleys were found to be filled 
with a concentrate of the more resistant constituents of the 
drift, the soluble constituents having apparently been removed 
just as in the weathered portion of the Kansan drift. The 
valleys are shallower than in the Kansan drift of southern 
Iowa and northeastern Missouri, and run out at their heads in 
so-called sloughs. But so far as discovered this shallowness is 
not occasioned by the deposition of a fresh sheet of drift, but 
rather seems due to a filling by slope wash. The ramification 
of the valleys into the uplands is not strikingly different from 
that found in the Kansan drift and seems to call for a maturity 
of stream work of a similar rank. In view of the fact that 
my own studies have not been exhaustive and that the Iowan 
invasion is still held by the Director of the Iowa Geological 
Survey, who has had a longer acquaintance with the region, I 
still consider the question of an Iowan invasion an open one. 

Ann Arbor, Michigan, February 22, 1909. 



C. H. Gordon — Chalk Formations of Northeast Texas. 369 



Art. XXIX. — The Chalk Formations of Northeast Texas /* 
by C. H. Gordon. 

It has long been known that deposits of chalk occur in 
northeast Texas, but writers are not in accord as to the relations 
of these formations. In a paper read before the Geological 
Society of America by Mr. R. T. Hill in 1893,f which was 
the successor to a number of papers on the Cretaceous of the 
Arkansas-Texas region by this author during the preceding 
six or seven years, mention was made of the occurrence of 
chalk in the southern part of Lamar county, and its identity 
w T ith the chalk of the White Cliffs of Arkansas was asserted. 
This chalk, which Mr. Hill states is "called the AnonaJ chalk 
in Texas," is regarded by him as distinct from the Austin chalk 
of central Texas and as belonging to a higher horizon. He says: 
" It is not known what has become of the Austin chalk in this 
section, but my hypothesis, backed by some evidence, is that to 
the southward it has been faulted down. The Anona (White 
Cliffs) chalk is an entirely distinct and higher bed, for it is 
underlain by the Taylor (Exogyra ponderosa) marls which 
overlie the Austin chalk." It is stated that the Austin chalk 
reappears in its normal position to the eastward, at Rocky 
Comfort in Little River county, Arkansas, on the north side 
of the Red River. 

In a paper on the chalks of southwestern Arkansas, published 
in the 22d Annual Report of the U. S. Geological Survey 
(p. 698), Mr. J. C. Taff refers to the Texas deposits and says : 
" The white chalk is exposed from Austin northward to Sher- 
man, Texas, through a distance of nearly 250 miles, without 
appreciable change in its thickness of nearly 600 feet, and with 
very slight variation in texture, color, and nature of material. 
* * * In the vicinity of' Sherman, a few miles south of Red 
River, the chalk formation turns eastward and continues down 
the south side of the valley through Grayson, Fannin, Lamar, 
and Red River counties, and into the northwestern part of Bowie 
county. From the last locality the chalk passes beneath the 
bottom land of the Red River to Rocky Comfort, Arkansas. 
Farther east it comes to the surface at White Cliffs, on Little 

* Read before the Geological Society of America at the Baltimore meeting, 
1908. Published by permission of the Director of the United States Geo- 
logical Survey, During a part of the time devoted to field study in north- 
east Texas, the author was assisted by Mr. L. F. Russ of the University of 
Texas. 

f Bulletin Geological Society of America, vol. v, pp. 297-338, 1893. 

^Hill's spelling of "Anona" was corrected by Veatch to Annona. See 
Professional Paper No. 46, U. S. Geological Survey, p. 25 (footnote). 



370 C. II. Gordon — Chalk formations of Northeast Texas. 

Kiver, and Saline Landing on the West Saline River. In the 
Arkansas region, it is known as the White Cliffs formation." 

In a recent paper* Veatch follows Hill in assigning the 
Annona chalk to a higher horizon than the Austin. The 
statement is madef that the Austin is not found east of Paris, 
Texas. In his table of correlations, the upper part of the 
Bingen sand is regarded as the equivalent of the Austin chalk. 
He says : " The Bingen, which is the lithological counterpart 
of the Woodbine, apparently contains littoral equivalents of 
the Austin and Eagle Ford." 

The formations underlying northeast Texas consist chiefly of 
Upper Cretaceous and lower Tertiary rocks. The Woodbine 
sands, which constitute the base of the Upper Cretaceous, 
appear along the south side of Bed Biver, in the northern part 
of Lamar county. The strata have a general dip of about 50 
to 60 feet per mile toward the south by southeast. As a result 
of this, in passing from north to south, higher formations 
appear in successive belts except where interrupted by cover- 
ings of more recent deposits. The classification of the Upper 
Cretaceous of northeast Texas as adopted in a forthcoming 
report on the underground waters of the region is as follows : 

Navarro and Taylor. 

6. Dark laminated clays and blue sandy shales ; limited expo- 
sures in the north part of Hopkins county. 

5. Glauconitic sands in the south part of Delta county. 

4. Sandy marls and clays with lime concretions filled with 
fossils. Underlies most of Delta county. 

3. Impure sandy chalk forty to fifty feet thick grading into marls 
above and the sandy beds below. Exposed in the north part of 
Delta county near Enloe and extends in a narrow belt southwest- 
ward somewhat to the south of Ladonia and between Fairlie 
and Wolf City in Hunt county. 

2. Fine yellow glauconitic sand with thin lenses of fossilifer- 
ous limestone. Grades upward into the chalk and downward 
into clay. Thickness about 40 to 60 feet. 

1. Blue clay marls forming a deep black soil similar to that of 
the marls above. One hundred to one hundred and fifty feet. 

Austin Formation. 

Annona Chalk. — Bluish and creamy-white chalk. Thickness 

fifty to one hundred feet. 
Brownstown Marls. — Blue marly clay, slightly glauconitic. Four 

hundred to five hundred feet thick. 

* Professional Paper, No. 46, U. S. Geological Survey, 1906, 
f Ibid, p. 19. See also p. 24. 



0. IT. Gordon — Chalk Formations of Northeast Texas. 371 

Eagle Ford Formation. 

Blossom Sands. — Glauconitic sands interlaminated with and 
grading into clay. Sands appear decidedly brown and red at 
the surface from oxidation of the contained iron. Eighty 
to one hundred feet thick. Constitute a narrow sandy belt 
extending from east to west through Red River and Lamar 
counties. Named from the village of Blossom in the eastern 
part of Lamar county, which is located upon their outcrop. 
Dark blue and black laminated clays with thinly bedded 
clay limestone and nodular septaria of blue limestone. 
These septaria are highly fossiliferoas and often weather 
out in bowlders of various sizes. These clays grade down- 
ward into the Woodbine sands and upward into the Blos- 
som sands. Their thickness is from five hundred to six 
hundred feet. 

Woodbine Formation. 

Ferruginous and argillaceous sands, in places glauconitic and 
lignitiferous. Thickness estimated to be five hundred to eight 
hundred feet. 

Steatigraphic Relations of the Annona Chalk. 

The chalk to which Hill gave the name Annona makes its 
appearance on the south side of Red River in the northern 
part of Red River county and extends in a belt 2 to 4 miles 
wide to the south of west across Red River and Lamar coun- 
ties, and westward. Owing to the hardness and greater resist- 
ance to erosion, this belt appears usually as a low ridge, 
covered to a slight depth only by soil through which the 
white chalk frequently appears, and in which ribbons of white 
mark the courses of wet-weather streams. The chalk con- 
tains a varying proportion of sand. 

In composition, the chalk corresponds closely with that of 
the Austin, as shown by the following analyses : 

l 

Calcium carbonate 84*14 

Silica and insolubles. 10-14 

Ferric oxide and alumina 4-04 

Magnesia (and H 2 0) 1'68 

1. Annona chalk, 7 miles south of Paris, Texas. Analysis fur- 

nished by Mr. J. A. Porter, Paris, Texas. 

2. Texas chalk, locality unknown. 21st Annual Report, U. S. 

Geological Survey,* Pt. VII, 1899-1900, p. 329. 

3. Rocky Comfort, Arkansas, Ibid. 

4. Average rock from quarry, Texas Portland Cement Works, 3 

miles west of Dallas, Texas, Ibid., p. 737. 



2 


3 


4 


82-51 


84-48 


70-21 


11-45 


9-77 


23-55 


3-61 


1-15 


1-50 


1-19 


trace 


0-58 



372 C. H. Gordon — Chalk Formations of Northeast Texas. 

Upon the outcrop of the chalk are located the towns of 
Clarksville, Atlas, Roxton, Honey Grove, etc. Annona, the 
town in Red River county from which the formation received 
its name, lies several miles south of the outcrop of the beds. 

About three miles west of Clarksville, the exposures of 
chalk and accompanying marls are interrupted by a covering 
of Quaternary deposits, partly filling a broad shallow valley 
excavated in the formation now occupied by the headwater 
branches of Cuthand Creek. 

At Clarksville several wells have been extended through 
chalk and underlying marls to the Blossom sands. No distinc- 
tion is made in the well records between the marl and the 
chalk, both being included under " white rock," which is said 
to be about 600 feet thick. 

The outcrop of the blue chalk marls, called the Brownstown 
marls, which underlie the Annona chalk, constitutes a belt of 
black land, several miles wide, along the north side of the 
chalk and bounded on the north by a strip of sands (Blossom 
sands) representing the top of the Eagleford formation. The 
thickness of the Brownstown in Red River and Lamar coun- 
ties is estimated at 400 to 500 feet. It thins toward the west, 
with a corresponding diminution in the width of outcrop, as 
shown in Fannin county. Accompanying this change in the 
thickness of the marls, there is an increase in thickness of the 
chalks, the north boundary of which passes to the northward 
of Honey Grove and thence to Sherman. 

The chalk beds occurring in the vicinity of Sherman have 
generally been recognized as Austin. It is evident therefore 
that the Annona chalk and the underlying Brownstown marls 
are stratigraphically the equivalents of the Austin chalk. 
The meager collections of fossils thus far obtained from these 
formations in northeastern Texas offer little in the way of 
faunal evidence bearing on the classification of these beds. 
The collections thus far made indicate a wide range of the 
species represented and no satisfactory conclusion can be 
drawn from them. 

One hundred to two hundred feet above the Annona chalk 
is another bed of chalk having a thickness of fifty to one 
hundred feet. This chalk outcrops along the south side of 
North Fork in the northern part of Delta county from the 
vicinity of Enloe to the western boundary of the county. 
The position of this chalk suggests its correlation with the 
Saratoga chalk in Arkansas, which occupies a similar position 
with respect to the Annona (White Cliffs) in that region. 
There are no intervening exposures of chalk, however, and 
the data from well borings seems to indicate that these upper 



C. R. Gordon — Chalk Formations of Northeast Texas. 373 

chalk formations occur in lenses rather than as persistent 
formations. 

Conclusions. 

1. The stratigraphic relations of the Annona chalk shows 
conclusively that it is the equivalent of the upper part of the 
Austin chalk formation of central Texas. 

2. That the Brownstown marl, which underlies the Annona 
in northeast Texas and western Arkansas, is the stratigraphical 
equivalent of the lower portion of the Austin. 

3. That the term Austin should be used for the formation 
as a whole, while the names Annona and Brownstown may be 
used to designate the chalk and marl members respectively. 

4. For the sands at the top of the Eagle Ford formation, 
which are of importance in certain areas as a source of water, 
the name Blossom is here proposed. The Clarksville wells 
draw their supply of water from these beds, in view of which, 
in the absence of knowledge of the outcrop of the formation, 
Yeatch called them the Sub-Clarksville Sands. 

Knoxville, Term. 



374 D. K. Greger — Devonian of Central Missouri. 



Art. XXX. — The Devonian of Central Missouri; by 
Darling K. Greger. 

The Devonian of central Missouri is a two-fold formation of 
somewhat variable thickness and in the region of its greatest 
development it consists of more or less heavy-bedded, blue, 
gray and reddish-brown limestones, and of coarse, blue and gray 
shales and shaly limestones, the latter weathering to a light 
} 7 ellow, highly siliceous clay. 

The occurrence of Devonian strata in the central Missouri 
region w r as known to Owen (1) in the early fifties, and he out- 
lined its area with some degree of accuracy as follows : "In 
Missouri this formation (Devonian) was traced, reappearing 
for a very limited space in the valley of the Auxvasses in 
Calloway county; skirting for a short distance one of the south- 
ern promontories of the Iowa and Missouri coal-field, in close 
proximity to the great uplift of magnesian limestone, of Silu- 
rian date, in the same vicinity. It has, probably, a consider- 
ably greater range in this locality than here ascertained and 
laid down by me." 

Numerous exposures of the strata here considered occur in 
Warren, Montgomery, Callaway, Boone, Cole, Moniteau and 
Cooper counties, with their maximum development in central 
Callaway county. While the Devonian of this region is readily 
separable into two divisions upon both faunal and lithologic 
grounds, the lower member only bears a distinctive name, and 
this even is ill-defined as it appears in the literature. 

C. R. Keyes (2), in a discussion of the geological formations 
of Missouri, defines the Callaway limestone or lower member 
in the following manner : "In southeastern Missouri, rocks 
containing the typical fauna of the Western Hamilton are 
sparingly represented in Perry and Cape Girardeau counties, 
in connection with the limestones above mentioned (Grand 
Tower limestone). In this region the limestones belonging to 
this group are dark colored shaly rocks, quite different from 
the associated strata. 

"JSTorth of the Ozark uplift the Devonian rocks referred to 
the Hamilton extend westward along the Missouri river as far 
as Jefferson City, having their most typical development in 
Callaway county. In several places abundant fossils of this 
formation have been obtained from strata having lithological 
characters not very unlike the beds of the eastern Ozark region 
referred to the same age." 

The two paragraphs above quoted are somewhat misleading, 
since if the writer means to apply the name Callaway lime- 






D. K. (rreger — Devonian of Central Missouri. 375 

stone to the strata "having their most typical development in 
Callaway county" he would likewise include beds in Perry and 
Cape Girardeau counties, which to our knowledge contain an 
entirely distinct assemblage of fossils, a fauna in fact intimately 
related to the southern Illinois Devonian, which has been 
shown by Weller (3), Schuchert (4), and Savage (5), to belong 
to the eastern or Mississippian Province. 

While we regret that the name Callaway limestone has 
assumed a place in our literature through a somewhat illegiti- 
mate course, its retention should be insisted upon since the 
formation has its greatest development both in area and thick- 
ness in Callaway county. 

To the upper member of the central Missouri Devonian we 
would apply the name Craghead Creek shale since the typical 
exposure of this division is found on Craghead Creek (6), a 
small tributary of Middle River, south of Fulton about six 
miles, ou the old Snyder farm. At this place about 35 feet of 
shale is exposed and the contact with the Mississippian above 
and the Callaway limestone below are both visible. The 
discovery of these shales dates back to 1858 and it is quite 
possible that Dr. S. S. Laws, then president of Westminster 
College, was the first collector to visit the locality, since it was 
through him that Prof. Swallow became acquainted with its 
rich fauna. 

A number of small outcrops of the Craghead Creek shale 
occur in Callaway county, but it is nowhere so fully exposed as 
on the Snyder farm, where within a distance of one mile the 
following section is exhibited: Lower Coal-measures, Upper 
Burlington, Lower Burlington, Kinderhook, Craghead Creek 
shale, Callaway limestone, first Magnesian limestone, Sacchar- 
oidal sandstone. 

Structurally the Craghead Creek shale is readily divisible 
into three members: the lower portion made up of dark blue 
and drab shale, with interbedded bands of shaly limestone, 
highly fossilif erous ; the middle member consisting of a light 
drab, argillaceous limestone with few fossils, mostly remains 
of Crinoidea ; the upper member consisting of a light gray, 
siliceous shale, this like the lower member being highly 
fossiliferous. 

With the exception of the Brachiopoda and Crinoidea the 
fossils of the Craghead Creek shale have received only a 
cursory study, hence the following notes on the fauna are almost 
exclusively confined to these forms. 

The faunal lists here presented are based on collections made 
in the following localities. Callaway limestone: Bellama 
Springs, southeast of Fulton, Auxvasse Church, east of 
McCredie, and Sampson's Ford, east of Fulton. The Craghead 



376 D. K. Greger — Devonian of Central Missouri. 

Creek shale list is that of the type locality. The names 
preceded by an asterisk are species common to the Lime 
Creek shale or the Cedar Valley limestone of Iowa. 

(A) = abundant, (C) = common, (R) = rare and (VR) = very 
rare. 

Fauna of the Craghead Creek Shale. 

Anthozoa. 

*Acervularia profunda Hall. (R) 
* Aulopora sp. ? (C) 
*Aidopora sp. ? (C) 
Heliophyllum ? sp. ? (C) 

Bryozoa. 

*Leioclema occidens White and Whitfield. (A) 

Crinoidea. 

Aristocrinus concavus Rowley. (VR) 
Melocrinus gregeri Rowley. (R) 
Melocrinus lylii Rowley. (R) 
Melocrinus tersus Rowley. (R) 

The four species of Crinoidea described by Prof. Rowley 
seem to be confined to the limestone constituting the middle 
member of the shale and have not been found outside of 
Callaway county. 

Brachiopoda. 

Athyris minima Swallow. (C) 

Atrypa gregeri Rowley. (C) 
*Atrypa hystrix oceidentalis Hall. (R) 
'""Atrypa reticularis Linnaeus. (A) 

* Camarotoechia sp. nov. (R) 
Cranaena morsii Greger. (R) 

* Cyrtina triquetra Hall. (C) 
*Dielasma calvini Hall and AVhitfield. (C) 

Dielasma sp. nov. (R) 

* Gypidula comis Owen. (R) 
Hypothyris sp. nov. (R) 
Leiorhynchus sp. nov. (R) 

* Philhedra famelica Hall and Whitfield. (C) 
Philhedra crenistria Hall. (R) 
Productella callawayensis Swallow. (C) 
Productella marquessi Rowley. (C) 

*Pugnax altus Calvin. (R) 

* Schizophoria iowensis Hall. (C) 






D. K. Greger — Devonian of Central Missouri. 377 

Schizophoria macfarlani Meek. (A) 

Schuchertella pandora Billings. (R) 

Spirifer amarus Swallow. (A) 
*Spirifer asper Hall. (R) 
*Spirifer euryteines Owen. (R) 

Stropheodonta sp. ? (A)f 

Strophonella crassa Rowley. (VR) 

Pelecypoda. 

Six species of this class are recognized in the fauna, but all 
are preserved only as casts of the interior and their relationship 
has not been determined. 

Gastropoda. 

The Gastropods, like the Pelecypods, are poorly preserved 
casts and their affinities are unknown. 

Cephalopoda. 

Nautilus lawsi Swallow. (R) 
Orthoceras sp. ? (C) 

Fauna of the Callaway Limestone. 
Porifera. 

* Stromotopora sp. ? (A) 

Anthozoa. 

* Acervidaria davidsoni Edwards and Haime. (A) 

Aulopora sp. ? (R) 
*Favosites alpenensis Winchell. (A) 
* Pachyphyllum woodmani White. (R) 

Zaphrentis, two indet. sp. (A) 

Crinoidea. 

Megistocrinus cf. latus Hall. (R) 
Melocrinus cf. nodosus Hall. (R) 

Brachiopoda. 

* Athyris fultonensis Swallow. (A) 
*Atrypa reticularis Linnaeus. (A) 

* Cranaen a iowensis Calvin. (C) 

fProf. Swallow in 1860 described in the St. Louis Academy of Science 
Transactions ten species of Stropheodonta from the Callaway limestone and 
Craghead Creek shale, without illustrations. The types were lost at the 
time the State University burned and from the descriptions specific identi- 
fication is impossible. 



37$ D. K. G'reger — Devonian of Central Missouri. 

Cyrtia occidentalis Swallow. (R) 

Cyrtina missouriensis Swallow. (A) 

Hypothyris sp. no v. (R) 
* Pentamerella dubia Hall. (R) 
* Philhedra famelica Hall and Whitfield. (R) 

ISpirifer annae Swallow. (A) 
*Spirifer iowaensis Owen. (R) 
* Stropheodonta costata Owen. (C) 

Strop heodonta sp. indet. (A) 

Pelecypoda. 

Conocardium sp. nov. (A) 

Gastropoda. 

Callonema ? sp. ? (R) 

JSuomphalus sp. ? (A) 

Loxonema ? sp. ? (C) 

Pleuro torn aria providensis Broadhead. (R) 

Eeperences. 

(1) Owen D. D., Geological Survey of Wisconsin, Iowa and Minnesota, 
1852, page 81. 

(2) Keyes, C. E., Geological Survey of Missouri, 1894, vol. iv, page 43. 

(3) Weller, S., Jour. Geol., vol. v, no. 6, 1897. 

(4) Schuchert, Charles, Amer. Geol., vol. xxxii, 1903. 

(5) Savage, T. E., this Journal, vol. xxv, 1908. 

(6) IT. S. Top. Atlas, Fulton Sheet, edition of 1890. 

Fulton, Missouri. 



Browning and Palmer — Estimation of Thallium. 379 



Art. XXXI. — The Volumetric and Gravimetric Estimation 
of Thallium in Alkaline Solution by Means of Potassium 
Ferricyanide ; by Philip E. Browning and Howard E. 
Palmer. 

[Contributions from the Kent Chemical Laboratory of Yale Univ. — cxcviii.] 

In a former paper from this laboratory* it was shown that 
reactions involving oxidation by ferricyanide in alkaline solu- 
tion and reoxidation by permanganate of the resulting ferro- 
cyanide in acid solution might be applied to the determination 
of cerium in the presence of the other rare earth compounds. 

In the work to be described the same reactions were applied 
to the volumetric estimation of thallium. A solution of thal- 
lium nitrate was made up by dissolving 10 grms. of pure 
thallous nitrate in water and making up to a liter. This solu- 
tion was standardized by precipitating definite portions with 
potassium bichromate in alkaline solution, filtering, and weigh- 
ing the thallous chromate,f and also by evaporating measured 
portions of the solution with an excess of sulphuric acid and 
weighing as the neutral sulphate ;J the mean of closely agreeing 
results by both methods was taken as the standard. 

Volumetric. — The procedure was as follows: Measured 
portions of the solution of thallous nitrate were drawn from a 
burette, water was added to about 100 cm3 a sufficient quantity of 
a solution of potassium ferricyanide to give an excess, and potas- 
sium hydroxide to complete precipitation of the brown thallic 
hydroxide. The precipitate was filtered off on asbestos, gener- 
ally without settling, and washed thoroughly. The filtrate was 
acidified with sulphuric acid and titrated with standard perman- 
ganate. From the following equations, representing the reac- 
tions, the amount of thallium present may be readily calculated: 

Tl 2 + 4K 3 FeC 6 N 6 + 4KOH --= Tl 2 3 + 4K 4 FeC 6 N 6 + 2H 2 
5K 4 FeC 6 N 6 + KMn0 4 + 4H 2 S0 4 = 

5K 3 FeC 6 N 6 + 3K 2 S0 4 + MnS0 4 + 4H 2 

It was found necessary to apply a correction for the amount of 
permanganate used to give the first tinge of pink color to the 
amounts of ferricyanide used in the determinations. This 
seldom exceeded '1 of l cm3 of the permanganate. 

The following table shows the results obtained with different 
amounts of the thallium salt: 

* Browning and Palmer, this Journal (4), xxvi, 83. 

f This Journal (4), viii, 460. % This Journal (4), ix, 137. 



380 Browning and Palme?' — Estimation of Thallium. 





T1,0 taken 


T1 2 


Error 




as the 


found 






nitrate 








grm. 


grm. 


grm. 


(1) 


0-0809 


0-0809 


+ 0-0000 


(2) 


0-0809 


0-0808 


—o-oooi 


(3) 


0-0809 


0-0809 


±0-0000 


(4) 


0-1213 


0-1212 


— 0-0001 


(5) 


0-1213 


0-1216 


+ 0-0003 


(6) 


0-1213 


0-1218 


+ 0-0005 


0) 


0-1213 


0-1218 


+ 0-0005 


(8) 


0-1213 


0-1212 


— 0-0001 


(9) 


0-1213 


0-1207 


— 0-0006 


(10) 


0-1618 


0-1614 


— 0-0004 


(H) 


0-1618 


0-1613 


-0-0005 


(12) 


0*1618 


0-1616 


— 0-0002 



Gravimetric. — The satisfactory character of the precipitate 
of thallic hydroxide obtained by this process suggested the 
possibility of applying it to a gravimetric estimation of thal- 
lium. Several precipitations were made according to the 
method described and the precipitate contained in a perforated 
platinum crucible upon an asbestos felt was dried over a 
low flame at about 200° C. to constant weight. The results 
follow in the table : 





T1 2 3 


T1 2 3 


Error 




taken 


found 






as the 








nitrate 








grm. 


grm. 


grm. 


(1) 


0-1305 


0:1309 


0-0004 + 


(2) 


0-1305 


0-1314 


0-0009 + 


(3) 


0-1305 


0*1308 


0*0003 + 


« 


0-0870 


0-0872 


0-0002 + 


(*) 


0-1740 


0-1741 


0-0001 + 


(6) 


0-1740 


0-1739 


o-oooi — 


(?) 


0-1740 


0*1742 


0-0002 + 


(8) 


0-1305 


0-1307 


0-0002 + 


(9) 


0-1305 


0-1309 


0-0004 + 


(10) 


01305 


0-1308 


0-0003 + 


(11) 


0-0870 


0-0872 


0-0002 + 


(12) 


0-0870 


0-0874 


0-0004 + 



In experiments (5) to (12) the precipitated thallic hydroxide 
was washed with hot water, a procedure to be recommended 
when large amounts are present. 



T D. A. Cockerel! — Descriptions of Tertiary Insects. 381 



Art. XXXII. — Descriptions of Tertiary Insects, VI ; by 

T. D. A. CoCKERELL. 

A Peculiar Neuropteroid Insect from Colorado. 
Eomerope gen. nov. (Meropidse.) 

Rather stout-bodied, with the end of the abdomen formed 
somewhat as in Dittacus, or nearer, perhaps, to the type of 
certain Trichoptera ; legs with many very strong spines ; wings 
shaped much as in Merope, but more elongate, and at the same 
time with the costal region more strongly arched, and includ- 
ing (between the costa and subcosta) four longitudinal series 
of cells. 

Eomerope tortriciformis sp. n. 

$ . General aspect that of a tortricid moth ; length 13 mm ; 
head small ; width of thorax about 4 ram , of abdomen about 3-J-, 
both no doubt somewhat widened by crushing ; color of body, 
as preserved, pale ferruginous, the apex of the abdomen darker, 
and a dark chitinous plate (? ventral) in the region of the basal 
half of the first abdominal segment (fig. 3) ; end of abdomen 
not elongated, but bearing a pair of thumb-shaped obtuse 
harpes, directed obliquely upwards (fig. 2) ; the uncus or corre- 
sponding structure short and apparently obtuse, without an 
upturned point ; legs with short femora (hind femur about 3 mm ), 
but rather long and slender tibiae and tarsi ; (hind tibia about 
5 mm ), the tibiae (at least) with delicate appressed hair 2 and 
numerous very large spines (fig. 1) which are finely striated 
longitudinally (fig. la). The structure of the tarsi, the antennae, 
mouth-parts, etc., cannot be seen. 

Wings as preserved pale reddish, nearly the color of the 
shale, faintly striated by the slightly darker veins but without 
spots or bands ; anterior wing 15-| mm long, and about 4 broad 
(the exact breadth difficult to determine), with the basal part of 
the costa strongly arched, but the apical two-thirds nearly 
straight (the costal outline much as in Apochrysa) ; apex 
rounded. Hind wings about 13 mm long and 3 broad, more 
slender than the anterior. 

Venation of anterior wings. — Subcosta remote from the 
costal margin, leaving a large space in which there are for a 
considerable distance four longitudinal rows of cells. The first 
upward branch bounds basally two large cells only, but the 
second branch curves forward and runs parallel with the main 
stem of the subcosta, and between it and the costal margin are 
three rows of cells, the second and third of the middle row 
being hexagonal (this is shown, better than it can be described, 
in fig. 5) ; toward the apex of the wing the symmetry is some- 

Am. Jour. Sci.— Fourth Series, Vol. XXVII, No. 161.— May, 1909. 
26 



382 T. D. A. Cockerell — Descriptions of Tertiary Insects. 

what lost and the cells become reduced, as shown in fig. 7. 
The cells of the second row from costa, after the third, be- 
come longer and lose their hexagonal form, having the lower 
side straight ; after the eighth cell the middle row is lost, 
there being two large cells in place of three, but there is a 

Figs. 1—5. 




Eomerope tortriciformis Ckll. 

Fig. 1. Part of anterior tibia, showing the fine hairs and large spines, 
la. part of spine magnified, showing the finely striated surface. 

Fig. 2. Apex of abdomen. 

Fig. 3. Chitinous plate in region of first abdominal segment. 

Fig. 4. Venation of anterior wing, showing relationship of radius, 
media, etc. 

Fig. 5. Venation of costal region of anterior wing near base. 

small middle cell immediately after, as shown in fig. 7. The 
area between the subcosta and its upper branch (they ulti- 
mately unite again as shown in fig. 7) is divided into three 
very long cells, the cross-veins being opposite the fourth of the 
second series of cells (shown in fig. 5) and the basal part of the 
seventh of the same series. From the subcosta to the radius 
are four cross-veins, placed at wide intervals, the fourth being 
shown in fig. 7 {Panorpa has usually two such cross-veins). 
The radius and media at first run side by side, so that upon 
superficial examination they look like one very stout vein ; at 
the point where the radius branches the media bends down- 
wards, as shown in fig. 4 (this differs from Panorpa in the 
earlier branching of the radius). The radial sector forks, and 
the lower branch forks again ; the upper branch of the latter 
fork again branches, and its upper branch is connected by a 
cross-vein with the upper main branch of the sector, so that 
there is enclosed an elongated cell, very acute basally, obliquely 



T. D. A. Cocker ell — Descriptions of Tertiary Insects. 383 

truncate apically, and with its lower side divided into three 
sections. The media and cubitus appear to be formed much as in 
Panorpa, and are connected by cross-veins, as shown in fig. 4. 
Their distal parts are obscured. There are two very strong 
but short curved anal veins, running downwards to the lower 
margin (much as in llerope, but shorter and more curved). 




Eomerope tortriciforniis Ckll. 

Fig. 6. Venation of hind wing, showing branches of radia. sector. 
Fig. 7. Venation of costal region of anterior wing toward apex. 

Venation of hind icings. — This cannot be wholly made out, 
but fig. 6 shows the discal region. The branches of the radial 
sector enclose a long cell much as in the anterior wing. 

Hah. — Miocene shales of Florissant, 1907 ; doubtless from 
Station 14, but the specimen is not marked with any number 
or collector's initial. Holotype in Yale University Museum. 

This is one of the most puzzling fossil insects I have had 
occasion to describe, but I believe it is correctly assigned to 
the Mecaptera. The form of the wing, with the strongly con- 
vex costa and numerous costal cells, is very different from that 
of Panorpa, Bittacits, etc., but the isolated and peculiar 
Merope, of the eastern United States, shows an approach to 
this condition, the costal area being broad, and some of the 
cells divided into two. On the other hand, the apical struc- 
tures of the abdomen in Eomerope do not resemble those of 
Merope, but are even more simple than Bittacus, showing 
resemblance to the doubtless more primitive condition found 
in the Trichoptera. There is no particular resemblance to the 
Mesozoic Orthophlebidse. 

liandlirsch divides the modern Mecaptera or Panorpatse into 
four families : Bittacusidse, Panorpidse, Meropidse, Boreiclae. 



384 T. D. A. Cockerell — Descriptions of Tertiary Insects. 

If this is admitted, apparently Meropidse may be divided into 
Meropinee, for Merope, and Eomeropinse, n. subf., for Eome- 
rope, the latter being separated on the structure of the abdomen 
and the venation of the anterior wings. It may be supposed 
that these insects represent an American type, once prevalent, 
but now reduced to a single genus and species, Merope tuber, 
Newman. 

A New Type of Mecaptera, approaching the JVemopteridas. 

JEobanksia bittaciformis gen. et sp. nov. 

Wings long and narrow, hyaline, with the apical half of the 
costa fuscous ; anterior wing about 16 mm long and 3 wide ; hind 
wing about 12 mm long and 2 wide. 

FjG g Anterior wing. — Costal 

margin of basal half of wing 
heavy, faintly arched, with 
several straight or very 
slightly oblique cross-nerv- 
ures to subcosta. On the 
apical half of the wing the 
narrow area between the ra- 
dius and costa is wholly fus- 
cous, and no cells are visible. 
Subcosta straight, running 
parallel with and close to the 
radius, its exact termination 
not visible on account of the 
darkening. Radius perfectly 
Fig. 8. Eobanksia bittaciformis straight, even toward the 
Ckll. base of the wing only half 

a millimeter from costal mar- 
gin, and terminating on costa a short distance before the 
obtuse apex of wing. Radial sector leaving radius about 6 
from base of wing, and branching after a course of about 3 
running practically as in Bittacus ; the lower branch simple ; 
the upper running near the radius, and forked after a course 
of 4 mm ; in another specimen the lower branch of this fork is 
again forked. Media appressed to radius at base, and after 3 mm 
leaving it at a very acute angle, and having a straight un- 
branched course, practically parallel with the stem and lower 
branch of the radial sector. Cubitus small and weak, running 
close to lower margin, and bending abruptly downwards to end 
upon it, after about 4 mm . Cross-nervures irregular and vari- 
able, but more numerous than the Bittacus, especially in the 
basal half of the wing. No anal visible, nor indeed is there 
room for one. 

Posterior wing. — Narrow, with two strong parallel closely 
adjacent straight veins running its entire length, about two- 



\ 



mm 
mm 



T. D. A. Cockerell — Descriptions of Tertiary Insects. 385 

fifths of the wing above, and three-fifths below them ; media 
represented by a straight rather weak vein in the lower field ; 
cross-nervures oblique, their more distal ends on the margin. 
The type consists of a pair of wings ; another specimen 
(Station 13 B, Miss Gertrude Darling) shows the thorax and 

Fig. 9. 




Fig. 9. Eobanksia bittaciformis Ckll. 

A. Diagram of wing of type. 

B. Branching of radial sector in another specimen. 

abdomen ; the head is missing. The thorax is about 6 mm long, 
the abdomen about 13, the latter curved, and apparently formed 
exactly as in Bittacus, although no details can be seen. When 
the wings are folded backwards, the costa is downward. 

Hah. — Miocene shales of Florissant, Colorado. A cotype in 
Yale Museum. This remarkable insect seems to form a new 
family (Eobanksiidse) of Mecaptera. The anterior wing is 
very like that of the Panorpids in many respects ; the hind 
wing closely resembles that of the Nemopteridge, though not 
so elongated, and without an apical expansion. (Some Nemop- 
teridee have no expansion.) Mr. JS". Banks, to whom I sent a 
sketch of the venation, agrees that there is a distinct approxi- 
mation to Nernoptera. Formerly it used to be maintained that 
the Panorpids (Mecaptera) and Nemopterids were allied ; the 
insect now described appears to lend support to this opinion. 
The genus is dedicated to Mr. !N". Banks, in recognition of his 
labors on the JNeuroptera. 

Trichoptera from Florissant. 

Phenacopsyche vexans gen. et sp. no v. 

Anterior wing. — Length about 19 mm (16-J visible, the base 
hidden), outer margin about 10 mm , lower margin about 14. 
Costa with basal half straight, apical slightly arched ; apex 
obtuse, outer margin regularly convex. Wing brownish, from 
a rather dense scaling or pubescence, the apical part of the 
costal region and the broad (3 ram ) outer margin darker than the 
rest (possibly less denuded) ; veins dark brown. 



386 



T D. A. Cockerell — Descriptions of Tertiary Insects. 



The wing lias an extraordinarily Lepidopterous appearance, 
being very much as in the broader-winged Noctuids, and even 
more like some of the Hesperiids, especially in the distinct anal 
angle. Regarding it as Lepidopterous, the visible portion of 
the venation nearly agrees with that of Ado?ieta, except for the 



Fig. 10. 



Fig. 11. 




Figs. 10 and 11. Phedacopsyehe vexans Ckll. 

insuperable difficulty of two extra veins between the supposed 
media and cubitus. . Treated as Trichopterous, this difficulty 
disappears, the supposed median cell being the discoidal, in the 
forks of the radial sector. 

The generic characters ascertained are as follows : — 
Discoidal cell present, elongate triangular ; no median cell ; 
radius (R x ) joined to first branch of radial sector (R 3 ) by 
a cross-nervure, about l mm (on R 2 ) beyond discoidal cell : at 
this cross-nervure (which is regularly present in Panovpa, 
and also occurs in various Trichoptera, as Rhyacophila, 
Odontoccrum, and Namamyia) the radius is bent a little 
upwards, as shown in the figure. The five branches of the 
radius, and four of the media, are all present, and consequently 
the apical cellules 1 to 7, none being either absent or stalked. 
R 2 and R 3 leave the discoidal cell close together near its apex, 
and R. leaves its lower corner, being in a line with its lower 
side. M 3 and M 4 are stronger than M L and M 2 , and appear to 
form the principal branches of the media, M 1 and M 2 leaving 
from an evanescent vein passing up from the media to the 
radius. Mr. K. Banks, to whom I sent a sketch of the vena- 
tion, remarks on the apparent absence of the anterior main 
branch of the media, which should go to M 1 and M 2 . I have 
carefully examined the specimen, and although the main stem 
of the media is strong and dark, I find no sign of this branch. 
It seems as if it had migrated forwards, to form the vein leav- 
ing M 3 for the bases of M, and M 2 , the latter being connected 
at the base also (by a cross-vein) with the radial sector. 



T. D. A. CocJcerell — Descriptions of Tertiary Insects. 387 

Apparently the genus must be referred to the Odontoceridse, 
but Mr. Banks intimates that this is hardly a natural family, 
but "a sort of waste-basket" for things hardly fitting into Seri- 

Fig. 12. 




Fig. 12. Phenacopsyche vexans Ckll. Diagram of venation. 

costomatidse, and without the necessary characters of Leptoce- 
riclse, Rhyacophilidse, or Hydropsychidse. 

Hob. — Miocene shales of Florissant, 1908 (Station 13 B, 
Geo. N. Roliwer). 

Hydropsyche scudderi sp. nov. 

Larva in general similar to that of Hydropsyche sp., figured 
in Bull. 47, New York State Museum (1901), plate 15, f. 3, but 
much longer (length about 31 mm ), though not broader (width 
of thorax about 3 mm ), and with the thoracic plates much more 
nearly equal in size, being, in lateral view, about as deep as long. 
Head and thoracic plates strongly chitinized, of the usual 
rather dark reddish brown color ; head 3 mm long and 2 deep, 
apparently quite normal ; abdomen visible only as a faintly 
darker shade, of about the same width as the thorax, having 
slight indications of dark transverse and longitudinal markings. 
On one specimen indications of the branched gills, yellowish 
in color, can be seen on the first three segments. The caudal 
end is slender and produced, the caudal legs provided with the 
usual spreading bunches of hair. No indication of any case. 

There are two good specimens before me, one showing the 
dorsal, the other the lateral aspect. Many others have been 
found ; those in which the abdomen cannot be clearly seen are 
curiously similar to Scudder's Planocephcdus aselloides. 

Hob. — Miocene shales of Florissant. 

Scudder is no doubt correct in assuming that the Hydropsy- 
chiclse of Lake Florissant did not breed in the lake, but in the 
small streams running into it. During the volcanic eruptions 
these streams may have been so heated that the larvae were 
killed and then washed into the lake. 

University of Colorado, Dec. 2, 1908. 



388 A\ F. Griggs — Divided Lakes in Western Minnesota. 



Art. XXXIII. — Divided Lakes in Western Minnesota ; by 
Robert F. Griggs. 

No feature of the physiography of western Minnesota is 
more notable to the casual observer than the large number of 
twin lakes. Throughout Becker and Ottertail counties and 
presumably over a much larger area a very large proportion of 
the lakes are partially or completely divided by narrow necks 
stretched across from shore to shore. This feature is very 
striking to the traveler in the region, because these narrow 
ridges cutting the lakes in two are almost always traversed by 
the main roads, which thus connect the country lying on oppo- 
site sides of the lake, the importance of the road depending on 
the size of the lake and the consequent amount of country 
benefited by the short cut. These dividing ridges are ice- 
pushed ramparts, features which, on lake margins, are well 
known and have been fully described by others. But no one, 
so far as the writer has found, has shown the possibilities of ice- 
shove in building new shore lines and cutting up the lakes in 
which it works. 

The mechanism of ice action and its effects in building ram- 
parts of sand and bowlders along the shore have been most fully 
described by Bulkley ('00) and Gilbert ('90, '08). Briefly it 
is this : If the ice covering a lake is subjected to further 
declines in temperature after freezing, it contracts and is broken 
up by cracks running in every direction ; water from below now 
seeps into these cracks and freezes. Thus the shrinkage in 
volume is taken up by new ice, so that when the temperature 
rises and the ice expands the whole sheet most enlarge. This 
enlargement causes the ice, carrying with it whatever bowlders 
or other material it may grip, to shove up onto the shore if it 
be low, or if it be a resistant cliff, to pile up against it, or if the 
lake be long and narrow, to buckle up in the open water where 
the major cracks due to lateral friction and other causes create 
zones of weakness. The shove may be very slight in the begin- 
ning but as it is repeated with every change of the weather 
until the ice is broken up, its cumulative effects are so great as 
to push, up large ramparts of bowlders or to cause considerable 
destruction along the shore. 

A typical example of an arm of a lake cut off from the main 
body by an ice-pushed rampart occurs near the northeastern 
angle of Lake Pelican (fig. 1), where a small pond is separated 
from the lake b} T such a barrier. This rampart, though very 
narrow, stands quite high above the water and supports a row 
of trees. Its back side slopes somewhat gently into the swampy 



H. F. Griggs — Divided Lakes in Western Minnesota. 389 

bottom of the pond. But its front slope (fig. 2) exhibits the 
typical features of an ice rampart. It is built of gravel with 
occasional bowlders of large size. It is very steep and in places 
the ice has shoved up over the bowlder front and cut into the 
turf of the top. 

Detroit Lake on the line of the Northern Pacific Railway is 
the most accessible example of a large lake with such barriers. 
In this case the process of division is not complete and the 
two ramparts form a pair of slender points stretching out from 
the opposite shores. A view of the southern point is shown in 

Fig. 1. 




Fig. 1. A dividing rampart, Lake Pelican. 

figure 3. It is remarkable in its extreme narrowness. Though 
nearly a mile long, it is hardly more than one hundred feet 
wide at the widest point and the tip is so sharp that the 
extremity is hardly wide enough for the feet of the observer. 
The opposite point reaching out from the northern shore is 
very similar but is not so long. The two are connected by a 
submerged bar over which the water is so shallow that boats 
can cross only in a narrow passage in the center. But there is 
no channel across and the passage is simply the lowest point in 
the continuous ridge. In character this point is very different 
from such ramparts as that shown in figure 1. It is built of 
sand and gravel from which bowlders are almost absent. It is 
nowhere high or steep-sided like the bowlder rampart, but rises 
scarcely three feet above water level. 

In many respects this point resembles sand spits formed by 
water work. The fine materials of which it is built and its 



390 B. F. Griggs — Divided Lakes in Western Minnesota. 

extremely sharp point both resemble such spits. Its structure 
and position show plainly, however, that water can have had 
practically no part in its construction. There is nothing of the 
Y-shaped ridging characteristic of cusps built out directly from 
the shoreline, nor are there any hooks such as are often built 
by long shore currents. Moreover this lake has no currents or 
waves competent to build such a spit. Its longest diameter is 
less than three miles and there are no currents along its shores 
except those due to local winds. Such waves as are raised by 
the prevailing westerly winds are directly opposed to the 



Fig. 2. 







Fig. 2. Front slope of rampart shown in fig. 1. 

building up of such north and south spits and, especially at 
their extremities, tend to destroy them and scatter the materials 
of the sharp ridge. 

That water currents can have no part in the formation of 
these spits is shown very clearly by another small lake some 
miles west of Detroit. This lake (tig. 4) is completely divided 
by one barrier while two others project from its shores. It is 
so small that water work of any magnitude is impossible. 
Indeed it is so small as to raise the question whether ice-push 
could become sufficiently great to heap up the ramparts 
observed. Such action is undoubtedly greatly favored in this 
lake by the very slight depth which would enable the ice to 
seize abundant' materials for its work. It is known that ice- 
shove is entirely absent from deep lakes, but to just what degree 
extreme shallowness would favor the contraction and expansion 
needs to be determined by careful field studies. 



R. F. Griggs— Divided Lakes in Western Minnesota. 391 

Even a superficial examination is sufficient to show that ice 
is the only agent now at work in the formation of these divid- 
ing barriers. It is easy to see that the effects of ice action on 
such submerged ridges as that described in Lake Detroit would 
accentuate the ridge and pile up the rampart till it emerged 
from the water and formed a complete barrier across the lake. 
It is not difficult to see how such a ridge would be formed on 
any shoal which the ice could reach and grapple, especially if 
it happened to be located, as is very often the case, along lines 

Fig. 3. 




Fig. 3. Partially completed dividing rampart. Lake Detroit. 

of the major cracks in the ice which tend to run from head- 
land to headland because of the strains set up by the independ- 
ent expansion and contraction of the separate bays of the lake. 
But it is difficult to understand how ramparts could be built 
out from the shore even along the major cracks where the ice 
buckles up and forms high ice ridges, unless there were a shal- 
low place in the beginning. But though these lakes are shallow 
as lakes go, areas shallow enough to be reached even by the very 
thick ice of the Minnesota winter are by no means abundant. 
Though it be demonstrated that ice is responsible for accentuat- 
ing and extending these ridges after they reach a certain point 
in their history, it is a question how much of their formation 
may be assigned to this agency and how much must be attributed 
to the original irregularities of the lakes, such as lines of moraine 
crossing them. This question can only be answered by extended 
field study in the region conducted during the winter, when the 
ice may be seen at work and its actual effects determined. 

But while it may be difficult to explain the division of large 
lakes completely by ice action, it is evident that the cutting off 
of small ponds and swamps from the main body of the lake is 



392 7?. F. Griggs — Divided Lakes in Western Minnesota. 

entirely explicable on this basis. Though such cases are less 
striking than the larger ramparts such as that in Lake Detroit, 
they are very much more numerous. Almost every large lake 
in the region has several sloughs cut off by ice ramparts. 

The effect of the formation of ice barriers across the bays 
of a lake is very marked on the later history of the lake. The 
primary effect is to dredge out the shallows to clean out the 

Fig. 4. 




Fig. 4. Small lake with three ramparts. 

advancing marsh vegetation and to pile up steep banks, thus 
enlarging the lake and counteracting the most potent agencies 
tending to fill it up. But when the barriers are thrown across 
so as to divide the lake into small parts, secondary effects 
appear which reverse the primary. As the reaches of water 
over which the wind sweeps are reduced, the consequent erosive 
action of waves on the shores is lessened and transportation 
by long shore currents is checked. This greatly accelerates 
the filling up of the lake by favoring the growth of vegetation 
in its shallow water. This is so marked that it is difficult to 
find such a bay which is not already far along toward extinc- 
tion. Most of them are so choked with vegetation that though 
their character is plainly evident on inspection it cannot be 
made to appear in a photograph such as figure 1. It is obvious 
that in regions where its action is well developed ice-shove is 
to be reckoned as an indirect aggradational factor of first impor- 
tance. 

Literature cited. 

Bulkley, E. R., Ice Ramparts. Trans. Wis. Acad. Soi., Arts and Letters, 
xii, pp. 141-157, 1900. 

Gilbert, G. K., Lake Bonneville. U. S. G. S. Monogr. 1, p. 57, 1890. 
, Lake Ramparts. Bull. Sierra Club, vi, pp. 225-234, 1908. 

Columbus, Ohio, Feb., 1909. 



Mixter — Heat of Formation of Titanium, Dioxide. 393 



Art. XXXIY. — The Heat of Formation of Titanium 
Dioxide, and third paper on the Heat of Combination of 
Acidic Oxides icith Sodium Oxide • by W. G. Mixter. 

[Contributions from the Sheffield Chemical Laboratory of Yale University] 

The thermochemistry of titanium is limited to that of the 
reaction TiCl 4 +Aq=57,870 c (Thomsen). As the heat of for- 
mation of the tetrachloride is not known, that of the oxide or 
acid can not be calculated. Good results might be obtained 
by burning titanium in oxygen if it could be ground to a fine 
powder. A coarse powder burns well in sodium peroxide, and 
hence this method was used.* 

The metal used in the investigation was made by Dr. M. A. 
Hunter in the laboratory of the General Electric Company. 
The specimen was in irregular nuggets, was malleable when 
hot and brittle when cold. The density of a bright piece of 
the metal was found to be 4*51 and that of the pulverized 
portion 4*49 at 18°. Moissanf states that the density of tita- 
nium is 4*87. But the purest he made in the electric furnace 
contained two per cent of carbon. 

For use in the combustions the metal was pulverized in 
diamond mortar and then ground in an agate one which it 
did not scratch. It was impossible to grind it to a fine powder, 
as some particles were flattened and others rounded. It was 
assumed that the impurities adhering to the nuggets were 
converted into dust by the grinding and also that the steel 
from the mortar was dust. Accordingly the dust was floated 
off in water and only the coarser portion retained. A qualita- 
tive analysis revealed no impurities in the metal. For a 
determination of titanium 0*5662 gram w r as dissolved in molten 
potassium pyrosulphate, the fusion w r as dissolved in cold water 
and the titanic acid was precipitated by boiling. The titanium 
dioxide obtained weighed 0*9453 gram, equivalent to 0*5676 
gram of titanium or 100*2 C per cent. The following are the 
thermal data : 

1 2 

Titanium _ 2*000 grams 2-001 grams 

Sulphur 1-000 " 1-000 " 

Sodium peroxide 20 " 20 " 

Water equivalent of system.. 3,003 " 2,979 " 

Temperature interval . _ 4*958° 4*958° 

*The details of the sodium-peroxide method are given in the first paper, 
on the Heat of Combination of Acidic Oxides with Sodium Oxide, in this 
Journal, xxvi, 125. 

fC. E., cxx, 290. 



394 Mixter — Heat of Formation of Titanium Dioxide. 

Heat effect.. 14,8S9 C 14,7Y0 C 

" of oxidation of sulphur. —5,27 l c —5,271° 

" " " " iron 6 — 48° 6 — 48 c 

" " oxygen absorbed — 7l c — 70° 

9,499 c 9,381° 

For 1 gram of titanium 4,749 c 4,693° 

The mean is 4,721° and for 48*1 grams of titanium it is 
227,100°. The temperature of the fusion was above the melt- 
ing point of silver and no metallic titanium remained. The 
fusion, when treated with water, left a white residue which 
dissolved on addition of an excess of hydrochloric acid and the 
solution had the red color of pertitanic acid. 

Titanic Oxide. 

Titanic oxide was prepared as follows : The hydroxide, 
purchased for the pure compound, was fused with potassium 
pyrosulphate, the fusion was dissolved in cold water and the 
titanic acid was precipitated by boiling. It was washed with hot 
water, digested with hot ammonia and washed again and 
finally heated to redness for an hour. It contained no iron, 
alumina or zirconia. The following are the experimental data : 

3 4 5 

Titanic oxide 3*302 gr. 4-001 gr. 3*970 gr. 

" " in residue, __ 0*057 " 0*074 " 0*103 " 

" " combined _.. 3*245 " 3*927 " 3*867 " 

Sulphur 1*000 " 1*000 " 1*000 " 

Sodium peroxide 16 " 18 " 19 " 

Water equivalent of system 3,081 "2,956 " 3,022 " 

Temperature interval 2*417° 2*580° 2*536° 

Heat effect 7,446° 7,626° 7,664° 

" of oxidation of sulph'r —5,271° —5,271° —5,277° 

" " " " iron.. —48° —48° —48° 

" " oxygen evolved. __ +78° +72° 

2,127° 2,385° 2,417° 

For 1 gram of titanic oxide 655° 607° 625° 

The mean is 629° and for 80 grains of titanic oxide it is 
50,300°. The solutions of the fusions in hydrochloric acid had 
the red color of pertitanic acid. 

Since a pertitanate is formed when titanium or its dioxide is 
fused with an excess of sodium peroxide, it is probable that the 



Mixtev — Beat of Formation of Titanium Dioxide. 395 

same compound results in both cases. Moreover, it makes no 
difference in the heat effect of Ti + 20 derived whether we 
assume the formation of Na 2 2 .Ti0 3 obtained by Mellikoff 
and Pessarjewsky* or JSTa 2 O.Ti0 3 . Taking the latter com- 
pound, the calculations are as follows : 



3Xa 2 2 + Ti = Na 2 O.Ti0 3 + 2Na 2 + 227,100 c 

3Na 2 + 30 = 3Na 2 2 + . 58,200 c 



285,300 c 
Na 2 + Ti0 + O = Na o 0.Ti0 3 69,700 c 



Ti + 20 = Ti0 2 (amorphous) + 215,600 



Na 2 2 + Ti0 2 = Na 2 O.Ti0 3 + 50,300 c 

Na 2 + O = Na 2 2 + 19,400 c 

Na s O + Ti0 2 + O = Na 2 O.Ti0 3 + ._ 69,700 c 

The large heat of formation of titanium dioxide was to be 
expected because of the difficulty of reducing the oxide to the 
metal. The heat of combination of Ti0 2 or Ti0 3 with Na 2 
cannot be calculated from the experimental results, but it is 
less than that of silicon dioxide. 

Lead. 

Lead, in a finely divided form for the work, was obtained by 
scraping a revolving cylinder of the metal with a thin tool. 
The experimental data were as follows : 

Lead 20 grams 

" unburned 0*94 

" burned 19*06 

Sulphur 1 

Sodium peroxide 20 

Water equivalent of system 3,982 

Temperature interval 2*822° 

Heat observed ] 1 ,237 c 

" of oxidation of sulphur — 5,271 c 

" " " " iron — 48 c 

" " oxygen absorbed — 62 c 



5,856 c 

For 1 gram of lead .... 307 c 

" 206*9 grams 63,500° 

*Ber. d. deutsch. cliem. Gesell., zxxi, 953. 



396 Mixter — Heat of Formation of Titanium Dioxide. 

The fusion was yellow and it yielded to hot water a little 
sodium plumbate. The insoluble yellowish brown powder left 
by the water liberated chlorine from hydrochloric acid — a 
proof of the presence of peroxide. 

The heat of Pb.O = 503 is Thomsen's. Tscheltzow* found 
for PbO.O=12'l, hence Pb.20^62'4, which is used in the 
following calculation : 



2NaX> + Pb = Na PbO + Na,0 + .... 63,500 



2Na 2 + 20 = 2Na„0 2 + 38,800° 

2Na 2 4- 20 + Pb = Na 2 Pb0 3 + 102,300 c 

Pb + 20 = Pb0 2 + 62,400° 

2Na 2 + Pb0 2 = Na 2 Pb0 3 + 39,900° 

As the fusion in the foregoing experiment yielded a mixture 
of lead oxides, it seemed best to use lead dioxide in place of 
the metal. For this purpose the dioxide w x as made by acting 
on lead acetate with a hypochlorite and also by treating reel 
lead with nitric acid, but the products were not pure enough 
for the purpose. Accordingly dioxide was prepared by elec- 
trolyzing a saturated solution of lead nitrate in dilute nitric 
acid. It was washed, dried, pulverized, and washed again to 
remove adhering nitrate. Finally the lead dioxide was heated 
to 280° until the weight was constant, and as it still retained 
water the temperature was raised until some red lead formed on 
the bottom of the beaker containing it. The composition of 
the product was as follows : Pb0 2 , 93'7 ; PbO, 6*0 ; H 2 0, 0*3 per 
cent. The correction for lead oxide and water would be small 
and was not made in the following experiments : 



Lead dioxide 

Sulphur 

Sodium peroxide . _ - . 

Water equivalent of system 
Temperature interval 



Heat observed 1 3,733 



a 



a 





2 






3 






20. 




o-l 


•ams 20. 


<»' 


rams 


9- 






u 


1-5 




a 


25 






a 


15 




a 


4,018 






a 


4,090 




a 


3-418 


o 




2-658° 






13,733° 








10,871° 






— 10,542° 








-7,906° 






— 48° 








— 48° 






+ 151° 








+ 260° 






3,294° 


3,177° 





" " iron . 
oxygen evolved . . 



For 1 gram of lead dioxide . . 165° 1 59° 

The mean is 162° for 1 gram of lead dioxide and for 238-9 
grams it is 38,700°, which agrees well with the result of experi- 

*C. R., c, 1458. 



Mixter — Heat of Formation of Titanium Dioxide. 397 

rnent 1. The fusion of 2 when treated with hot water left 
lead oxide and but little dioxide, while in 3, where less sodium 
peroxide was used, considerable lead dioxide remained. These 
facts indicate that the lead dioxide which separates on the 
hydrolysis of sodium plumbate is reduced to oxide by sodium 
peroxide in the presence of water. The following results 
support this view : a mixture of the two oxides was placed in 
cold water and the whole was heated. The solution contained 
lead, but gave no chlorine after adding hydrochloric acid and 
warming. A fusion of equal parts of lead and sodium per- 
oxide left, after exhausting with water, lead dioxide, while when 
two parts sodium peroxide were taken the insoluble residue 
was brownish yellow and contained but little lead dioxide. 

The heat of hydrolysis of sodium plumbate is derived as fol- 
lows : 

Na 2 O.Pb0 2 + Aq = 2(Na.O.H.Aq) + Pb0 2 - Na 2 .O.Pb0 2 - H 2 .0 
15,500 c = 223,600 — 139,700 —68,400 

Since the Pb0 2 is in the solid state before and after the 
hydrolysis of the plumbate, it makes no difference in the ther- 
mal result if it first combines with water and is finally dehy- 
drated. The hydrolysis as given above is only complete in a 
large volume of water, since sodium plumbate is soluble as 
such in a concentrated solution of sodium hydroxide. 

Zirconium Dioxide. 

Two experiments were made with mixtures of zirconium 
dioxide, sodium peroxide, and sulphur. The fusions were not 
good, and the residues remaining after treatment with hot 
water set free chlorine from hydrochloric acid, indicating the 
presence of a peroxide. The results were 251 c and 268 c for 1 
gram of zirconium dioxide, mean 258 c and for 1227 grams 
31,700 c . If the dioxide was all oxidized to trioxide 19,400° are 
to be added, giving 51,100 c . The only interpretation to be given 
to the result is that the heat of Zr0 2 or Zr0 3 +Na 2 is small. 

Ceric Oxide. 

But one experiment was made with ceric oxide with the 
result of 94. 6 C for 1 gram and 16,300 c for 172 grams. The fusion 
was placed in cold water and after gas ceased to come off the 
solution was decanted. When the insoluble residue was treated 
with hot water much gas was evolved — an indication of the 
j)resence of a peroxide of cerium. If 2Ce0 2 is oxidized by 
sodium peroxide to Ce 2 5 9,700 c are to be added for the heat of 
the oxygen taken from sodium peroxide. The thermal result 
does not indicate combination of sodium oxide with an oxide 
of cerium. 

Am. Jour Sci.— Fourth Series, Vol. XXVII, No. 161.— May, 1909. 
27 



398 C. Palache — JVote on Crystal Form of Benitoite. 

Art. XXX Y. — Note on Crystal Form of Benitoite; by C. 

Palache. 

The symmetry class to which the crystals of the interesting 
new mineral benitoite belong conld not be definitely established 
by the forms hitherto observed upon them. Rogers* assigns it 
to one of two classes, the trigonal-bipyramidal (Class 19 of 
Groth) or the ditrigonal-bipyramidal (Class 22 of Groth) with 
probabilities favoring the latter. Neither of these classes has 
hitherto had a representative among crystals. 

Crystals of benitoite recently acquired by the Harvard 
Mineral Cabinet present a new form which establishes the 
correctness of Rogers' assumption of its ditrigonal-bipyramidal 
symmetry. The form is the second order pyramid (2241); it 
occurs on all the crystals on the specimen with distinct but 
small faces which uniformly present a dull luster in striking 
contrast to that of the trigonal pyramids. The lateral edge 
between these second order pyramid faces is in most cases 
truncated by the second order prism with similarly dull faces. 
These two forms, characteristic for class 22, would appear as 
trigonal forms in class 19 ; the fact that when a face of either 
occurs, others are present of like quality and with six-fold 
repetition, seems to determine the type positively. Although 
dull the new faces gave distinct reflections on the goniometer. 
The average of six excellent readings, 0001 to 2211, was 71° 15'. 

The average of ten excellent readings, 0001 to 1011, was 
40° 12', a value just half way between those obtained by Rogers 
(40° 10') and Lauderbach (10° 14 7 ). Calculated from this value 
a : c=l : 0'7319 and ^> =^4879. Calculated from this axial 
ratio the angle 0001 to 2241 is 71° 10'. The list of forms for 
benitoite is then as follows: 

Prisms : m (1010), /* (0110), 
a (1120). 

Pinacoid : c (0001). 

Trigonal pyramids: p(1011), 
7T (0111), e (0112). j,. 

Second order pyramid : x 
(2241). j 

The figure shows a tj^pical 
combination. 

Octahedrite, a mineral not before recorded from the benitoite 
locality, was observed on part of the specimen from which 
natrolite' had been wholly removed by solution in hydrochloric 
acid. It appears in groups of pale brown crystals, combinations 
of unit pyramid and base ; the crystals are small and present 
facetted and curved faces so that they could not be measured 
but chemical tests showed the presence of titanic oxide alone. 

Harvard Mineralogical Laboratory, March, 1909. 
* Science, xxviii, 616, 1908. 




C. Palache and H. E. Merwin — Alamosite. 399 



Art. XXXYI. — Alamosite, a new Lead Silicate from Mexico ; 
by C. Palache and H. E. Merwin. 

In this paper is presented the description of a monoclinic 
lead metasilicate showing in form, habit, and composition close 
analogies with wollastonite, with which it is regarded as isomor- 
phous. This mineral was sent to the Harvard Mineralogical 
Laboratory for identification by the Foote Mineral Co. of 
Philadelphia, who generously placed at our disposal their 
whole supply of the material. According to their meager 
data regarding its occurrence it is found in an undeveloped 
gold and copper prospect situated near Alamos, Sonora, Mexico. 
The minerals making up the ore in hand are, however, with 
trifling exceptions all compounds of lead. 

The gangue is in part massive white quartz, in part a com- 
pact gray, brown or black material shown by analysis to con- 
tain quartz, the new lead silicate and either limonite or 
hematite, in varying admixtures. Interspersed through this 
massive material are occasional F ^ 

vugs lined with quartz crystals ^ ' 

and irregular bunches of the lead Vj ===± ====^ p 
compounds. Of these the most / v 

abundant is cerussite in snow- / \ ______ 

white aggregates and rare crys- / j 
tals. Minute flakes of pale green i 

leadhillite were identified by I J- a _ 

cleavage, optical character and 
chemical reactions. Wulfenite is [_ /__ 
also found, partly in orange-col- ' ~-^^^^-^^zzzj^g 

orecl crystals, more abundantly as 

a bright yellow stain in all the other minerals, of the ore, par- 
ticularly in the lead silicate. 

The lead silicate is in radiated fibrous aggregates of more 
or less pronounced spheroidal form, irregularly interspersed 
among the minerals already mentioned. It is snow-white in 
the mass, transparent and colorless in the rare cases where tiny 
fibers had been free to develop singly in open spaces between 
the spheroids. There is a perfect cleavage transverse to the 
fibers yielding a curved concentric fracture surface of pearly 
luster which extends almost uninterruptedly through all the 
individuals of a spheroid. The few developed crystals that 
could be secured for study were minute — not more than 0'5 mm 
in diameter — and the best of them were but poorly adapted 
to measurement, several fibers being generally adherent in 
subparallel groups so that it was difficult to secure readings 



400 C. Palache and II. E. Merwin — Alamosite. 

from individual crystals. The high luster of the mineral, 
however, made it possible to obtain readings from the most 
minute faces and fairly consistent measurements were finally 
obtained from six crystals. 

Alamosite is monoclinic, the fibers elongated parallel to the 
axis of symmetry (crystallographic axis b). This habit and 
the minute size of the crystals made it necessary to mount 
them on the two-circle goniometer with 010 as pole and the 
orthodome zone as prism ; the measurements and calculated 
angle§ of the table are, therefore, given for that position, but 
the symbols and axial ratio are for the normal position. The 
position was chosen so as to bring out as well as might be, the 
relation in form to wollastonite. 

The crystals are simple showing the forms <?(001), $(100), 
6(010), w&(110), tf(101), (7(011) i?(121), and r(121). The figure 
shows a typical combination in which the forms named, except 
m and r, are represented in about their normal development. 
The relations to wollastonite are shown by the following 
angles : 



Axial Ratio 



Alamosite Wollastonite 

a : b : c a : b : c 



1-375 : 1 : 0'924 1-053 : 1 : 0-967 

Angle 001 to" 100=0 84°10' 84°30' 

" 001 to 101 32 02 40 03 

" 001 to 011 42 36 43 55 

" 100 to 110 53 50 46 21 

Table of calculated and observed angles of alamosite, with 
010 as pole and 100 as first meridian. 

Elements for this position : ^ = 1*088 ; q = 0'731 ; / u=84°10' 

Elements for normal position : ^> =0 # 672; ^=0-919; fx=84 10 





Calculated 






Measured (m< 


san) 


No. o: 






i 


£ 




9 


< 


£ 


J 


o 




c 


001 


84' 


3 10' 


90 


D 00' 


84 ( 


D 10' 


90' 


w 


8 


a 


100 


00 


00 


90 


00 


00 


00 


90 


00 


6 


b 


010 


00 


00 


00 


00 


00 


00 


00 


00 


6 


m 


110 


00 


00 


36 


10 


00 


00 


36 


00 


1 


V 


101 


52 


08 


90 


00 


51 


53 


90 


00 


3 


9 


011 


84 


10 


47 


24 


84 


10 


47 


29 


6 


P 


121- 


-60 


11 


31 


57 


— 60 


12 


31 


58 


9 


r 


121 


52 


03 


34 


25 


52 


10 


34 


18 


2 



Physical Properties: — Cleavage perfect parallel to 010, 
therefore across the fibers. Specific gravity 6*488 ±'003, 
determined in the pycnometer on *6 CC of mineral (Merwin). 
Hardness 4-5. Luster adamantine. Plane of the optic axes 



C. Palache and II. E. Merwin — Alamosite. 401 

parallel to plane of symmetry. Refraction and double refrac- 
tion high but not determined. 

Chemical Composition : — A small amount of the mineral 
was picked out under the microscope as free as possible from 
adhering quartz and other substances contained in the ore. 
The absence of cerussite was proved by lack of effervescence 
in nitric acid. The mineral fuses at 3 to a greenish yellow 
bead, colorless when cold ; it is easily reduced on charcoal to 
a lead button and is soluble in nitric acid with strong o-elatini- 
zation. 

The analysis by H. E. Merwin gave the following result : 

Per cent 

Si0 o 21-11 

PbO 78*13 

CaO trace 

FeO . -09 

Residue* from PbO . '53 

Insol. residue, quartz -08 





Per cent 


Mol. Eatio 


for PbSiO: 


•348 


21-32 


•351 


78-68 



99'94 100-00 

* This residue was lost after weighing and was therefore not determined. 

The molecular ratio of PbO to Si0 2 is almost exactly 1 : 1, 
indicating for the mineral the formula PbSi0 3 . 

There seems little doubt after considering its characteristics 
that alamosite should be classified with wollastonite. That it 
is to be regarded as isomorphous with that mineral is, however, 
open to question. In favor of this interpretation are (1) sim- 
ilarity of chemical type and behavior with acids ; (2) identity 
of crystal system and habit ; (3) close approximation in the 
values of the angle /3 and in the lengths of the c-axes ; (4) 
similarity of optical orientation. Opposed to the assumption 
of isomorphism are (1) difference in lengths of the a-axes ; 
(2) difference of cleavage. 

The case is analogous to the relation of anglesite and anhy- 
drite where, however, isomorphism is less strongly indicated 
than in the present pair of minerals. 

Alamosite is very similar in appearance to barysilite but 
may be readily distinguished from it by its optical characters. 

Harvard University, March, 1909. 



402 C. Bar us — Absence of Polarization in Artificial Fogs. 

Art. XXXVII. — Absence of Polarisation hi Artificial Fogs ; 

by C. Barus. 

The astonishing feature of Tyndall's experiment, made in 
the usual way with motes of mastic suspended in water, is the 
completeness of the polarization of the light reflected or scat- 
tered at right angles to the impinging beam. This is particu- 
larly well shown with the double image prism, in which, for 
horizontal beams of light with vertical and horizontal vibration 
beams and line of sight horizontally at right angles to each other, 
only one of the beams (that for which vibration is vertical) is vis- 
ible in the turbid water. The other is quite extinguished while 
lying in a vertical plane either above or below the visible beam. 

In making the same experiment with dense fogs, I was 
surprised to find an almost entire absence of this discriminating 
character. Both beams are always in view and about equally 
intense. These fogs were produced in the fog chamber with 
phosphorus nuclei and were intensely luminous, with several 
millions of water particles per cubic centimeter of an average 
size less than 'OOOl 0111 . Even for fogs so dense and therefore 
so fine that the light penetrates scarcely 30 cm , both beams were 
about equally brilliant. On examining the two beams with a 
Kicol, however, they are found to be almost completely 
polarized at right angles to each other ; whence it follows that 
the vibration, which is horizontal outside, has been turned 
about 90 degrees in a horizontal plane after entering. In 
other words, whereas it vibrates in a horizontal plane normal 
to the observer on the outside of the fog chamber, it vibrates 
in a horizontal plane parallel to the observer in the inside of 
the fog chamber ; or, while the plane of the two vibrations is 
normal to the primary beam on the outside, it is parallel to the 
two beams on the inside of the fog chamber. The beam with 
vertical vibrations naturally remains unchanged. Finally, if 
coronas are produced from a polarized source, they are found 
to be polarized throughout all their colors and quite extin- 
guished between crossed JSficols. 

As the motes do not produce coronas and will not subside, 
it is difficult to specify their size. But the marked occurrence 
of scattering is sufficient evidence for the absence of regular 
reflection. Virtually at least, the particles are very small. 
With regard to the fogs, however, even in the case of the very 
finest particles, the light is still regularly reflected and refracted, 
and not scattered, as in the first instance, to an appreciable 
extent. Hence, for a line of vision normal to the primary 
beam, the direction of vibration may be turned 90 degrees 
parallel to the plane of the beam and vision, and now vibrates 
normal to the line of vision, seeing that light is not completely 
polarized on reflection from a surface of water. 

Brown University, Providence, R. I. 



Chemistry and Physics. 403 



SCIENTIFIC INTELLIGENCE. 



I. Chemistry and Physics. 

1. Prussian Blue and TurnbulPs Blue. — Many investigators 
have studied the well-known blue precipitates produced on the 
one hand by mixing solutions of